Upload folder using huggingface_hub

.gitattributes

@@ -139,3 +139,9 @@ checkpoints/codi_recon_delta_1.5b_ls1_rw0.03/checkpoint-300/tokenizer.json filte
 checkpoints/codi_recon_delta_1.5b_ls1_rw0.03/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/codi_recon_delta_1.5b_ls1_rw0.03/checkpoint-900/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/codi_recon_delta_1.5b_ls1_rw0.03/wandb/offline-run-20260629_145519-02eqg5ht/run-02eqg5ht.wandb filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-900/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/wandb/offline-run-20260629_034102-low3bm0v/run-low3bm0v.wandb filter=lfs diff=lfs merge=lfs -text

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/added_tokens.json

@@ -0,0 +1,35 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|action_sep|>": 151670,
+  "<|arg_sep|>": 151671,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|call_sep|>": 151666,
+  "<|end_of_text|>": 151673,
+  "<|endoftext|>": 151643,
+  "<|exception_sep|>": 151669,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|frame_sep|>": 151672,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|latent_end|>": 151675,
+  "<|latent_start|>": 151674,
+  "<|line_sep|>": 151667,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|return_sep|>": 151668,
+  "<|trace_context_start|>": 151665,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/config.json

@@ -0,0 +1,58 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8574b22bc2d1ae22b628c19daada22bdd27af0ccb445f58bee2976ee8eb9aa95
+size 4539424633

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b7395ec81531f7ea858214ac9a585d43b9ef8f195b3e3eb5ef21d65e49ba9a9
+size 3097074235

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4aa03284661460f7588cf2ed51cae6f25c4941e23d7976dea3db0885dadb428
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97017e17d6838d8a09f84b48e28a5938a238a3e50e6a6671b6a2439f52a40fa6
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5a6d9e5e8614e5b7bcfddbab25b719e621f1ff67f400bc9a73dc5af42659928
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:841e9f5079dd6c6b4131e29e2e887ebf784990e17a6d8d8c7131ffb11723d069
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ca01a26e4215a06318d34f4837223ef5e3e71cfd78483b4d43840df556602f8
+size 1465

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/thought_projector.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1d3c73d71253ae5c25908ca38c808d14d6c46b850723273ef45196830f272db
+size 9445953

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83a790d654474f5dfe225f889afd0210313eb1083f942671f2c4b8e95a1c922b
+size 11424004

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/tokenizer_config.json

@@ -0,0 +1,295 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<|trace_context_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151666": {
+      "content": "<|call_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151667": {
+      "content": "<|line_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151668": {
+      "content": "<|return_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151669": {
+      "content": "<|exception_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151670": {
+      "content": "<|action_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151671": {
+      "content": "<|arg_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151672": {
+      "content": "<|frame_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151673": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151674": {
+      "content": "<|latent_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151675": {
+      "content": "<|latent_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/trainer_state.json

@@ -0,0 +1,2914 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.34594594594594597,
+  "eval_steps": 500,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014414414414414415,
+      "grad_norm": 468.0,
+      "kd_loss": 0.22270642220973969,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 5.334,
+      "recon_loss": 6.499545574188232,
+      "recon_trunc": 0.0,
+      "step": 5,
+      "student_loss": 3.2635982036590576,
+      "teacher_loss": 1.4361417293548584
+    },
+    {
+      "epoch": 0.002882882882882883,
+      "grad_norm": 430.0,
+      "kd_loss": 0.21239201724529266,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 5.0745,
+      "recon_loss": 5.345489025115967,
+      "recon_trunc": 0.0,
+      "step": 10,
+      "student_loss": 2.728395462036133,
+      "teacher_loss": 1.1053788661956787
+    },
+    {
+      "epoch": 0.004324324324324324,
+      "grad_norm": 584.0,
+      "kd_loss": 0.18098419904708862,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 5.0678,
+      "recon_loss": 3.018347978591919,
+      "recon_trunc": 0.0,
+      "step": 15,
+      "student_loss": 1.5548276901245117,
+      "teacher_loss": 1.2046127319335938
+    },
+    {
+      "epoch": 0.005765765765765766,
+      "grad_norm": 246.0,
+      "kd_loss": 0.23498597741127014,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 5.1555,
+      "recon_loss": 2.2839651107788086,
+      "recon_trunc": 0.0,
+      "step": 20,
+      "student_loss": 2.183858871459961,
+      "teacher_loss": 0.6433351635932922
+    },
+    {
+      "epoch": 0.007207207207207207,
+      "grad_norm": 113.0,
+      "kd_loss": 0.2332078069448471,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 4.8142,
+      "recon_loss": 4.708770751953125,
+      "recon_trunc": 0.0,
+      "step": 25,
+      "student_loss": 2.056137800216675,
+      "teacher_loss": 1.7688019275665283
+    },
+    {
+      "epoch": 0.008648648648648649,
+      "grad_norm": 55.5,
+      "kd_loss": 0.21290458738803864,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 4.5809,
+      "recon_loss": 3.1192433834075928,
+      "recon_trunc": 0.0,
+      "step": 30,
+      "student_loss": 2.1787378787994385,
+      "teacher_loss": 1.8894840478897095
+    },
+    {
+      "epoch": 0.01009009009009009,
+      "grad_norm": 138.0,
+      "kd_loss": 0.168921560049057,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 4.3972,
+      "recon_loss": 6.197048664093018,
+      "recon_trunc": 0.0,
+      "step": 35,
+      "student_loss": 2.7991840839385986,
+      "teacher_loss": 1.9365880489349365
+    },
+    {
+      "epoch": 0.011531531531531532,
+      "grad_norm": 49.0,
+      "kd_loss": 0.1717594712972641,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 4.349,
+      "recon_loss": 3.7998454570770264,
+      "recon_trunc": 0.0,
+      "step": 40,
+      "student_loss": 2.2687275409698486,
+      "teacher_loss": 0.9082421660423279
+    },
+    {
+      "epoch": 0.012972972972972972,
+      "grad_norm": 63.75,
+      "kd_loss": 0.14987584948539734,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 4.092,
+      "recon_loss": 3.81005859375,
+      "recon_trunc": 0.0,
+      "step": 45,
+      "student_loss": 2.9089572429656982,
+      "teacher_loss": 2.3437206745147705
+    },
+    {
+      "epoch": 0.014414414414414415,
+      "grad_norm": 79.0,
+      "kd_loss": 0.1680448353290558,
+      "learning_rate": 9.999813520519569e-06,
+      "loss": 3.3431,
+      "recon_loss": 2.844719171524048,
+      "recon_trunc": 0.0,
+      "step": 50,
+      "student_loss": 1.6103482246398926,
+      "teacher_loss": 1.0730512142181396
+    },
+    {
+      "epoch": 0.015855855855855857,
+      "grad_norm": 35.0,
+      "kd_loss": 0.15395811200141907,
+      "learning_rate": 9.999055971469864e-06,
+      "loss": 3.1765,
+      "recon_loss": 1.9536027908325195,
+      "recon_trunc": 0.0,
+      "step": 55,
+      "student_loss": 1.1843489408493042,
+      "teacher_loss": 0.9362330436706543
+    },
+    {
+      "epoch": 0.017297297297297298,
+      "grad_norm": 34.75,
+      "kd_loss": 0.13685326278209686,
+      "learning_rate": 9.997715786107228e-06,
+      "loss": 2.9941,
+      "recon_loss": 1.636413812637329,
+      "recon_trunc": 0.0,
+      "step": 60,
+      "student_loss": 1.8569369316101074,
+      "teacher_loss": 0.6590576171875
+    },
+    {
+      "epoch": 0.018738738738738738,
+      "grad_norm": 35.0,
+      "kd_loss": 0.14608074724674225,
+      "learning_rate": 9.995793120629282e-06,
+      "loss": 2.8168,
+      "recon_loss": 1.7937917709350586,
+      "recon_trunc": 0.0,
+      "step": 65,
+      "student_loss": 1.148065447807312,
+      "teacher_loss": 0.3985089659690857
+    },
+    {
+      "epoch": 0.02018018018018018,
+      "grad_norm": 36.5,
+      "kd_loss": 0.09525534510612488,
+      "learning_rate": 9.993288199121283e-06,
+      "loss": 2.6115,
+      "recon_loss": 1.9947160482406616,
+      "recon_trunc": 0.0,
+      "step": 70,
+      "student_loss": 1.265057921409607,
+      "teacher_loss": 1.1560856103897095
+    },
+    {
+      "epoch": 0.021621621621621623,
+      "grad_norm": 30.75,
+      "kd_loss": 0.13154004514217377,
+      "learning_rate": 9.990201313529995e-06,
+      "loss": 2.3955,
+      "recon_loss": 1.9634184837341309,
+      "recon_trunc": 0.0,
+      "step": 75,
+      "student_loss": 0.8753498792648315,
+      "teacher_loss": 0.38040170073509216
+    },
+    {
+      "epoch": 0.023063063063063063,
+      "grad_norm": 31.5,
+      "kd_loss": 0.09402483701705933,
+      "learning_rate": 9.986532823629675e-06,
+      "loss": 2.0952,
+      "recon_loss": 1.7278754711151123,
+      "recon_trunc": 0.0,
+      "step": 80,
+      "student_loss": 1.5180137157440186,
+      "teacher_loss": 0.5871555805206299
+    },
+    {
+      "epoch": 0.024504504504504504,
+      "grad_norm": 30.75,
+      "kd_loss": 0.09141870588064194,
+      "learning_rate": 9.982283156980133e-06,
+      "loss": 1.8898,
+      "recon_loss": 3.0904629230499268,
+      "recon_trunc": 0.0,
+      "step": 85,
+      "student_loss": 1.0262964963912964,
+      "teacher_loss": 0.8668848276138306
+    },
+    {
+      "epoch": 0.025945945945945945,
+      "grad_norm": 46.25,
+      "kd_loss": 0.07249625027179718,
+      "learning_rate": 9.977452808876907e-06,
+      "loss": 1.6755,
+      "recon_loss": 2.737169027328491,
+      "recon_trunc": 0.0,
+      "step": 90,
+      "student_loss": 0.7890284061431885,
+      "teacher_loss": 0.6225405931472778
+    },
+    {
+      "epoch": 0.02738738738738739,
+      "grad_norm": 27.125,
+      "kd_loss": 0.058416809886693954,
+      "learning_rate": 9.972042342293531e-06,
+      "loss": 1.5593,
+      "recon_loss": 2.357276678085327,
+      "recon_trunc": 0.0,
+      "step": 95,
+      "student_loss": 0.6519027352333069,
+      "teacher_loss": 0.5244261026382446
+    },
+    {
+      "epoch": 0.02882882882882883,
+      "grad_norm": 59.25,
+      "kd_loss": 0.06716058403253555,
+      "learning_rate": 9.966052387815923e-06,
+      "loss": 1.3621,
+      "recon_loss": 1.56527841091156,
+      "recon_trunc": 0.0,
+      "step": 100,
+      "student_loss": 0.6716882586479187,
+      "teacher_loss": 0.4660281240940094
+    },
+    {
+      "epoch": 0.03027027027027027,
+      "grad_norm": 23.75,
+      "kd_loss": 0.06475857645273209,
+      "learning_rate": 9.959483643568893e-06,
+      "loss": 1.1764,
+      "recon_loss": 1.3081544637680054,
+      "recon_trunc": 0.0,
+      "step": 105,
+      "student_loss": 0.8320690393447876,
+      "teacher_loss": 0.5672342777252197
+    },
+    {
+      "epoch": 0.031711711711711714,
+      "grad_norm": 18.75,
+      "kd_loss": 0.08629752695560455,
+      "learning_rate": 9.95233687513477e-06,
+      "loss": 1.1259,
+      "recon_loss": 2.5731332302093506,
+      "recon_trunc": 0.0,
+      "step": 110,
+      "student_loss": 0.5341929197311401,
+      "teacher_loss": 0.2394336462020874
+    },
+    {
+      "epoch": 0.033153153153153155,
+      "grad_norm": 17.125,
+      "kd_loss": 0.05716023966670036,
+      "learning_rate": 9.944612915464183e-06,
+      "loss": 1.0453,
+      "recon_loss": 1.2201905250549316,
+      "recon_trunc": 0.0,
+      "step": 115,
+      "student_loss": 0.458853542804718,
+      "teacher_loss": 0.3588285446166992
+    },
+    {
+      "epoch": 0.034594594594594595,
+      "grad_norm": 13.3125,
+      "kd_loss": 0.04839833825826645,
+      "learning_rate": 9.936312664778978e-06,
+      "loss": 0.9859,
+      "recon_loss": 1.4293768405914307,
+      "recon_trunc": 0.0,
+      "step": 120,
+      "student_loss": 0.26777344942092896,
+      "teacher_loss": 0.23071233928203583
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 9.3125,
+      "kd_loss": 0.044529400765895844,
+      "learning_rate": 9.92743709046729e-06,
+      "loss": 0.9496,
+      "recon_loss": 2.0335958003997803,
+      "recon_trunc": 0.0,
+      "step": 125,
+      "student_loss": 0.5978028774261475,
+      "teacher_loss": 0.4314694404602051
+    },
+    {
+      "epoch": 0.037477477477477476,
+      "grad_norm": 10.875,
+      "kd_loss": 0.0389556884765625,
+      "learning_rate": 9.917987226970811e-06,
+      "loss": 0.8594,
+      "recon_loss": 4.102884769439697,
+      "recon_trunc": 0.0,
+      "step": 130,
+      "student_loss": 0.32229799032211304,
+      "teacher_loss": 0.2882220149040222
+    },
+    {
+      "epoch": 0.03891891891891892,
+      "grad_norm": 9.25,
+      "kd_loss": 0.0547921285033226,
+      "learning_rate": 9.907964175664207e-06,
+      "loss": 0.8408,
+      "recon_loss": 1.5782232284545898,
+      "recon_trunc": 0.0,
+      "step": 135,
+      "student_loss": 0.319209486246109,
+      "teacher_loss": 0.17503127455711365
+    },
+    {
+      "epoch": 0.04036036036036036,
+      "grad_norm": 9.625,
+      "kd_loss": 0.03733699396252632,
+      "learning_rate": 9.897369104726772e-06,
+      "loss": 0.8013,
+      "recon_loss": 3.04408597946167,
+      "recon_trunc": 0.0,
+      "step": 140,
+      "student_loss": 0.5711938738822937,
+      "teacher_loss": 0.4925714135169983
+    },
+    {
+      "epoch": 0.041801801801801805,
+      "grad_norm": 7.78125,
+      "kd_loss": 0.045995380729436874,
+      "learning_rate": 9.886203249006265e-06,
+      "loss": 0.8012,
+      "recon_loss": 0.9889851212501526,
+      "recon_trunc": 0.0,
+      "step": 145,
+      "student_loss": 0.16144226491451263,
+      "teacher_loss": 0.14944866299629211
+    },
+    {
+      "epoch": 0.043243243243243246,
+      "grad_norm": 7.875,
+      "kd_loss": 0.03740796819329262,
+      "learning_rate": 9.874467909874994e-06,
+      "loss": 0.7521,
+      "recon_loss": 1.144235372543335,
+      "recon_trunc": 0.0,
+      "step": 150,
+      "student_loss": 0.427932471036911,
+      "teacher_loss": 0.19212418794631958
+    },
+    {
+      "epoch": 0.044684684684684686,
+      "grad_norm": 8.625,
+      "kd_loss": 0.044521529227495193,
+      "learning_rate": 9.862164455078141e-06,
+      "loss": 0.7225,
+      "recon_loss": 0.9426382184028625,
+      "recon_trunc": 0.0,
+      "step": 155,
+      "student_loss": 0.29743173718452454,
+      "teacher_loss": 0.16907070577144623
+    },
+    {
+      "epoch": 0.04612612612612613,
+      "grad_norm": 9.1875,
+      "kd_loss": 0.049049314111471176,
+      "learning_rate": 9.849294318574353e-06,
+      "loss": 0.7009,
+      "recon_loss": 1.624841332435608,
+      "recon_trunc": 0.0,
+      "step": 160,
+      "student_loss": 0.293436199426651,
+      "teacher_loss": 0.10665447264909744
+    },
+    {
+      "epoch": 0.04756756756756757,
+      "grad_norm": 9.75,
+      "kd_loss": 0.034911200404167175,
+      "learning_rate": 9.835859000368615e-06,
+      "loss": 0.6697,
+      "recon_loss": 1.7004108428955078,
+      "recon_trunc": 0.0,
+      "step": 165,
+      "student_loss": 0.28085291385650635,
+      "teacher_loss": 0.18177205324172974
+    },
+    {
+      "epoch": 0.04900900900900901,
+      "grad_norm": 10.375,
+      "kd_loss": 0.04185783863067627,
+      "learning_rate": 9.821860066337427e-06,
+      "loss": 0.6457,
+      "recon_loss": 0.664492130279541,
+      "recon_trunc": 0.0,
+      "step": 170,
+      "student_loss": 0.17568205296993256,
+      "teacher_loss": 0.07977014034986496
+    },
+    {
+      "epoch": 0.05045045045045045,
+      "grad_norm": 7.96875,
+      "kd_loss": 0.03978830203413963,
+      "learning_rate": 9.807299148046301e-06,
+      "loss": 0.5758,
+      "recon_loss": 1.4172769784927368,
+      "recon_trunc": 0.1666666716337204,
+      "step": 175,
+      "student_loss": 0.35040274262428284,
+      "teacher_loss": 0.3842689096927643
+    },
+    {
+      "epoch": 0.05189189189189189,
+      "grad_norm": 9.5625,
+      "kd_loss": 0.03696564584970474,
+      "learning_rate": 9.792177942559598e-06,
+      "loss": 0.5569,
+      "recon_loss": 0.8903741240501404,
+      "recon_trunc": 0.0,
+      "step": 180,
+      "student_loss": 0.14949475228786469,
+      "teacher_loss": 0.07417529076337814
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 7.34375,
+      "kd_loss": 0.05137650668621063,
+      "learning_rate": 9.776498212242748e-06,
+      "loss": 0.4894,
+      "recon_loss": 0.9219234585762024,
+      "recon_trunc": 0.0,
+      "step": 185,
+      "student_loss": 0.23556068539619446,
+      "teacher_loss": 0.07139157503843307
+    },
+    {
+      "epoch": 0.05477477477477478,
+      "grad_norm": 7.3125,
+      "kd_loss": 0.04745712876319885,
+      "learning_rate": 9.76026178455684e-06,
+      "loss": 0.4877,
+      "recon_loss": 1.5784884691238403,
+      "recon_trunc": 0.0416666679084301,
+      "step": 190,
+      "student_loss": 0.12499237805604935,
+      "teacher_loss": 0.11093194782733917
+    },
+    {
+      "epoch": 0.05621621621621622,
+      "grad_norm": 6.78125,
+      "kd_loss": 0.026790278032422066,
+      "learning_rate": 9.743470551845625e-06,
+      "loss": 0.4296,
+      "recon_loss": 0.8518009781837463,
+      "recon_trunc": 0.0,
+      "step": 195,
+      "student_loss": 0.12112321704626083,
+      "teacher_loss": 0.07340381294488907
+    },
+    {
+      "epoch": 0.05765765765765766,
+      "grad_norm": 5.90625,
+      "kd_loss": 0.021310845389962196,
+      "learning_rate": 9.726126471114985e-06,
+      "loss": 0.3969,
+      "recon_loss": 1.7430474758148193,
+      "recon_trunc": 0.0,
+      "step": 200,
+      "student_loss": 0.18235476315021515,
+      "teacher_loss": 0.18640166521072388
+    },
+    {
+      "epoch": 0.0590990990990991,
+      "grad_norm": 6.875,
+      "kd_loss": 0.032883960753679276,
+      "learning_rate": 9.708231563804828e-06,
+      "loss": 0.3764,
+      "recon_loss": 0.8245623111724854,
+      "recon_trunc": 0.0,
+      "step": 205,
+      "student_loss": 0.042524177581071854,
+      "teacher_loss": 0.037807777523994446
+    },
+    {
+      "epoch": 0.06054054054054054,
+      "grad_norm": 6.34375,
+      "kd_loss": 0.04910533130168915,
+      "learning_rate": 9.689787915553497e-06,
+      "loss": 0.3654,
+      "recon_loss": 1.0637208223342896,
+      "recon_trunc": 0.021276595070958138,
+      "step": 210,
+      "student_loss": 0.12626872956752777,
+      "teacher_loss": 0.05190720781683922
+    },
+    {
+      "epoch": 0.06198198198198198,
+      "grad_norm": 5.53125,
+      "kd_loss": 0.030972279608249664,
+      "learning_rate": 9.670797675954693e-06,
+      "loss": 0.3442,
+      "recon_loss": 0.6881563663482666,
+      "recon_trunc": 0.0,
+      "step": 215,
+      "student_loss": 0.0411461777985096,
+      "teacher_loss": 0.033981483429670334
+    },
+    {
+      "epoch": 0.06342342342342343,
+      "grad_norm": 4.375,
+      "kd_loss": 0.028355753049254417,
+      "learning_rate": 9.651263058306932e-06,
+      "loss": 0.3024,
+      "recon_loss": 0.8817124366760254,
+      "recon_trunc": 0.0,
+      "step": 220,
+      "student_loss": 0.048761989921331406,
+      "teacher_loss": 0.048245809972286224
+    },
+    {
+      "epoch": 0.06486486486486487,
+      "grad_norm": 5.53125,
+      "kd_loss": 0.036966532468795776,
+      "learning_rate": 9.631186339355599e-06,
+      "loss": 0.2944,
+      "recon_loss": 0.8595229387283325,
+      "recon_trunc": 0.0,
+      "step": 225,
+      "student_loss": 0.05236447602510452,
+      "teacher_loss": 0.04269436374306679
+    },
+    {
+      "epoch": 0.06630630630630631,
+      "grad_norm": 5.8125,
+      "kd_loss": 0.022456347942352295,
+      "learning_rate": 9.610569859027585e-06,
+      "loss": 0.3147,
+      "recon_loss": 0.9198775887489319,
+      "recon_trunc": 0.0,
+      "step": 230,
+      "student_loss": 0.07517363876104355,
+      "teacher_loss": 0.07353860139846802
+    },
+    {
+      "epoch": 0.06774774774774775,
+      "grad_norm": 4.4375,
+      "kd_loss": 0.047675903886556625,
+      "learning_rate": 9.589416020158577e-06,
+      "loss": 0.3035,
+      "recon_loss": 0.5122413039207458,
+      "recon_trunc": 0.0,
+      "step": 235,
+      "student_loss": 0.04444430023431778,
+      "teacher_loss": 0.02076696790754795
+    },
+    {
+      "epoch": 0.06918918918918919,
+      "grad_norm": 6.46875,
+      "kd_loss": 0.03567289933562279,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.2806,
+      "recon_loss": 1.0555890798568726,
+      "recon_trunc": 0.0,
+      "step": 240,
+      "student_loss": 0.08203621208667755,
+      "teacher_loss": 0.039299674332141876
+    },
+    {
+      "epoch": 0.07063063063063063,
+      "grad_norm": 5.21875,
+      "kd_loss": 0.019655324518680573,
+      "learning_rate": 9.545506190996695e-06,
+      "loss": 0.2586,
+      "recon_loss": 0.7666318416595459,
+      "recon_trunc": 0.0,
+      "step": 245,
+      "student_loss": 0.05770008638501167,
+      "teacher_loss": 0.04188302904367447
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 4.40625,
+      "kd_loss": 0.03212125971913338,
+      "learning_rate": 9.52275531836226e-06,
+      "loss": 0.2374,
+      "recon_loss": 0.9074819684028625,
+      "recon_trunc": 0.0,
+      "step": 250,
+      "student_loss": 0.0864541307091713,
+      "teacher_loss": 0.08012009412050247
+    },
+    {
+      "epoch": 0.07351351351351351,
+      "grad_norm": 4.34375,
+      "kd_loss": 0.021410593762993813,
+      "learning_rate": 9.499477321907243e-06,
+      "loss": 0.2344,
+      "recon_loss": 1.2554527521133423,
+      "recon_trunc": 0.0,
+      "step": 255,
+      "student_loss": 0.026796536520123482,
+      "teacher_loss": 0.021643495187163353
+    },
+    {
+      "epoch": 0.07495495495495495,
+      "grad_norm": 3.78125,
+      "kd_loss": 0.0317341685295105,
+      "learning_rate": 9.475674914665084e-06,
+      "loss": 0.2393,
+      "recon_loss": 0.8790825009346008,
+      "recon_trunc": 0.0,
+      "step": 260,
+      "student_loss": 0.06856456398963928,
+      "teacher_loss": 0.026282260194420815
+    },
+    {
+      "epoch": 0.0763963963963964,
+      "grad_norm": 5.15625,
+      "kd_loss": 0.02957923151552677,
+      "learning_rate": 9.451350870788922e-06,
+      "loss": 0.2357,
+      "recon_loss": 0.9207150340080261,
+      "recon_trunc": 0.0,
+      "step": 265,
+      "student_loss": 0.04473090544342995,
+      "teacher_loss": 0.011104145087301731
+    },
+    {
+      "epoch": 0.07783783783783783,
+      "grad_norm": 5.71875,
+      "kd_loss": 0.01421283558011055,
+      "learning_rate": 9.426508025228255e-06,
+      "loss": 0.2302,
+      "recon_loss": 1.4175376892089844,
+      "recon_trunc": 0.0,
+      "step": 270,
+      "student_loss": 0.03786235302686691,
+      "teacher_loss": 0.0359700582921505
+    },
+    {
+      "epoch": 0.07927927927927927,
+      "grad_norm": 5.4375,
+      "kd_loss": 0.01542786043137312,
+      "learning_rate": 9.401149273398547e-06,
+      "loss": 0.2341,
+      "recon_loss": 0.6009160876274109,
+      "recon_trunc": 0.0,
+      "step": 275,
+      "student_loss": 0.016522981226444244,
+      "teacher_loss": 0.015627892687916756
+    },
+    {
+      "epoch": 0.08072072072072072,
+      "grad_norm": 4.40625,
+      "kd_loss": 0.043450385332107544,
+      "learning_rate": 9.37527757084375e-06,
+      "loss": 0.223,
+      "recon_loss": 1.9021402597427368,
+      "recon_trunc": 0.013513513840734959,
+      "step": 280,
+      "student_loss": 0.04809149354696274,
+      "teacher_loss": 0.023484718054533005
+    },
+    {
+      "epoch": 0.08216216216216216,
+      "grad_norm": 5.75,
+      "kd_loss": 0.020015748217701912,
+      "learning_rate": 9.348895932891857e-06,
+      "loss": 0.213,
+      "recon_loss": 1.0542433261871338,
+      "recon_trunc": 0.0,
+      "step": 285,
+      "student_loss": 0.07135617733001709,
+      "teacher_loss": 0.08110564947128296
+    },
+    {
+      "epoch": 0.08360360360360361,
+      "grad_norm": 4.15625,
+      "kd_loss": 0.01307689305394888,
+      "learning_rate": 9.322007434303449e-06,
+      "loss": 0.2256,
+      "recon_loss": 2.2934460639953613,
+      "recon_trunc": 0.0,
+      "step": 290,
+      "student_loss": 0.021568821743130684,
+      "teacher_loss": 0.020716898143291473
+    },
+    {
+      "epoch": 0.08504504504504505,
+      "grad_norm": 4.5,
+      "kd_loss": 0.03521936759352684,
+      "learning_rate": 9.29461520891335e-06,
+      "loss": 0.2203,
+      "recon_loss": 0.99751877784729,
+      "recon_trunc": 0.0,
+      "step": 295,
+      "student_loss": 0.05675162002444267,
+      "teacher_loss": 0.022469762712717056
+    },
+    {
+      "epoch": 0.08648648648648649,
+      "grad_norm": 6.0625,
+      "kd_loss": 0.02655652165412903,
+      "learning_rate": 9.266722449265371e-06,
+      "loss": 0.2058,
+      "recon_loss": 0.666858434677124,
+      "recon_trunc": 0.0,
+      "step": 300,
+      "student_loss": 0.11719609797000885,
+      "teacher_loss": 0.10709235072135925
+    },
+    {
+      "epoch": 0.08792792792792793,
+      "grad_norm": 3.359375,
+      "kd_loss": 0.029727192595601082,
+      "learning_rate": 9.238332406240224e-06,
+      "loss": 0.2164,
+      "recon_loss": 0.9334930181503296,
+      "recon_trunc": 0.0,
+      "step": 305,
+      "student_loss": 0.09936419874429703,
+      "teacher_loss": 0.07992291450500488
+    },
+    {
+      "epoch": 0.08936936936936937,
+      "grad_norm": 4.46875,
+      "kd_loss": 0.018055178225040436,
+      "learning_rate": 9.209448388676636e-06,
+      "loss": 0.2128,
+      "recon_loss": 0.8777124881744385,
+      "recon_trunc": 0.0,
+      "step": 310,
+      "student_loss": 0.009199777618050575,
+      "teacher_loss": 0.017623435705900192
+    },
+    {
+      "epoch": 0.09081081081081081,
+      "grad_norm": 2.953125,
+      "kd_loss": 0.018136998638510704,
+      "learning_rate": 9.180073762985701e-06,
+      "loss": 0.212,
+      "recon_loss": 3.670213460922241,
+      "recon_trunc": 0.0,
+      "step": 315,
+      "student_loss": 0.038779761642217636,
+      "teacher_loss": 0.03765730932354927
+    },
+    {
+      "epoch": 0.09225225225225225,
+      "grad_norm": 4.8125,
+      "kd_loss": 0.024612927809357643,
+      "learning_rate": 9.15021195275854e-06,
+      "loss": 0.1868,
+      "recon_loss": 0.6943050622940063,
+      "recon_trunc": 0.0,
+      "step": 320,
+      "student_loss": 0.023704711347818375,
+      "teacher_loss": 0.015942463651299477
+    },
+    {
+      "epoch": 0.0936936936936937,
+      "grad_norm": 3.390625,
+      "kd_loss": 0.028975514695048332,
+      "learning_rate": 9.119866438367263e-06,
+      "loss": 0.2102,
+      "recon_loss": 0.6595507264137268,
+      "recon_trunc": 0.0,
+      "step": 325,
+      "student_loss": 0.03306879103183746,
+      "teacher_loss": 0.01671084389090538
+    },
+    {
+      "epoch": 0.09513513513513513,
+      "grad_norm": 4.78125,
+      "kd_loss": 0.015513032674789429,
+      "learning_rate": 9.089040756559352e-06,
+      "loss": 0.2023,
+      "recon_loss": 1.7074724435806274,
+      "recon_trunc": 0.0,
+      "step": 330,
+      "student_loss": 0.045213621109724045,
+      "teacher_loss": 0.05958978086709976
+    },
+    {
+      "epoch": 0.09657657657657658,
+      "grad_norm": 3.265625,
+      "kd_loss": 0.015127932652831078,
+      "learning_rate": 9.057738500045453e-06,
+      "loss": 0.1969,
+      "recon_loss": 0.23923180997371674,
+      "recon_trunc": 0.0,
+      "step": 335,
+      "student_loss": 0.018725764006376266,
+      "teacher_loss": 0.013689985498785973
+    },
+    {
+      "epoch": 0.09801801801801802,
+      "grad_norm": 3.1875,
+      "kd_loss": 0.028886206448078156,
+      "learning_rate": 9.025963317080641e-06,
+      "loss": 0.1873,
+      "recon_loss": 0.8297758102416992,
+      "recon_trunc": 0.0,
+      "step": 340,
+      "student_loss": 0.013248264789581299,
+      "teacher_loss": 0.008833049796521664
+    },
+    {
+      "epoch": 0.09945945945945946,
+      "grad_norm": 3.25,
+      "kd_loss": 0.025645364075899124,
+      "learning_rate": 8.993718911039221e-06,
+      "loss": 0.1796,
+      "recon_loss": 0.6143622994422913,
+      "recon_trunc": 0.0,
+      "step": 345,
+      "student_loss": 0.009944843128323555,
+      "teacher_loss": 0.008340315893292427
+    },
+    {
+      "epoch": 0.1009009009009009,
+      "grad_norm": 7.8125,
+      "kd_loss": 0.0352986641228199,
+      "learning_rate": 8.96100903998311e-06,
+      "loss": 0.2098,
+      "recon_loss": 0.5921121835708618,
+      "recon_trunc": 0.0,
+      "step": 350,
+      "student_loss": 0.043480295687913895,
+      "teacher_loss": 0.013983724638819695
+    },
+    {
+      "epoch": 0.10234234234234234,
+      "grad_norm": 6.15625,
+      "kd_loss": 0.030174925923347473,
+      "learning_rate": 8.927837516223824e-06,
+      "loss": 0.2183,
+      "recon_loss": 0.5348331928253174,
+      "recon_trunc": 0.0,
+      "step": 355,
+      "student_loss": 0.02008548192679882,
+      "teacher_loss": 0.016475331038236618
+    },
+    {
+      "epoch": 0.10378378378378378,
+      "grad_norm": 4.0,
+      "kd_loss": 0.029513539746403694,
+      "learning_rate": 8.89420820587817e-06,
+      "loss": 0.2158,
+      "recon_loss": 0.8284482359886169,
+      "recon_trunc": 0.0,
+      "step": 360,
+      "student_loss": 0.06975475698709488,
+      "teacher_loss": 0.061798688024282455
+    },
+    {
+      "epoch": 0.10522522522522522,
+      "grad_norm": 3.203125,
+      "kd_loss": 0.039453133940696716,
+      "learning_rate": 8.86012502841764e-06,
+      "loss": 0.1834,
+      "recon_loss": 0.584143340587616,
+      "recon_trunc": 0.01785714365541935,
+      "step": 365,
+      "student_loss": 0.045608483254909515,
+      "teacher_loss": 0.019969221204519272
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 3.140625,
+      "kd_loss": 0.011938858777284622,
+      "learning_rate": 8.825591956211614e-06,
+      "loss": 0.1756,
+      "recon_loss": 0.805449366569519,
+      "recon_trunc": 0.0,
+      "step": 370,
+      "student_loss": 0.015335263684391975,
+      "teacher_loss": 0.016898587346076965
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 4.1875,
+      "kd_loss": 0.025258298963308334,
+      "learning_rate": 8.790613014064371e-06,
+      "loss": 0.1869,
+      "recon_loss": 0.6771914958953857,
+      "recon_trunc": 0.0,
+      "step": 375,
+      "student_loss": 0.020782332867383957,
+      "teacher_loss": 0.004855796694755554
+    },
+    {
+      "epoch": 0.10954954954954955,
+      "grad_norm": 3.359375,
+      "kd_loss": 0.024407049641013145,
+      "learning_rate": 8.755192278746e-06,
+      "loss": 0.1994,
+      "recon_loss": 0.9554619193077087,
+      "recon_trunc": 0.0,
+      "step": 380,
+      "student_loss": 0.14683908224105835,
+      "teacher_loss": 0.12159093469381332
+    },
+    {
+      "epoch": 0.110990990990991,
+      "grad_norm": 2.640625,
+      "kd_loss": 0.020979849621653557,
+      "learning_rate": 8.719333878517274e-06,
+      "loss": 0.209,
+      "recon_loss": 1.0486088991165161,
+      "recon_trunc": 0.0,
+      "step": 385,
+      "student_loss": 0.00612660963088274,
+      "teacher_loss": 0.005216435994952917
+    },
+    {
+      "epoch": 0.11243243243243244,
+      "grad_norm": 3.765625,
+      "kd_loss": 0.0173763670027256,
+      "learning_rate": 8.683041992648474e-06,
+      "loss": 0.1965,
+      "recon_loss": 0.13260489702224731,
+      "recon_trunc": 0.0,
+      "step": 390,
+      "student_loss": 0.012201618403196335,
+      "teacher_loss": 0.011553959921002388
+    },
+    {
+      "epoch": 0.11387387387387388,
+      "grad_norm": 3.09375,
+      "kd_loss": 0.0316915288567543,
+      "learning_rate": 8.646320850932326e-06,
+      "loss": 0.1632,
+      "recon_loss": 0.35131657123565674,
+      "recon_trunc": 0.0,
+      "step": 395,
+      "student_loss": 0.022094447165727615,
+      "teacher_loss": 0.005157520994544029
+    },
+    {
+      "epoch": 0.11531531531531532,
+      "grad_norm": 2.625,
+      "kd_loss": 0.026090312749147415,
+      "learning_rate": 8.609174733191012e-06,
+      "loss": 0.1835,
+      "recon_loss": 2.001973867416382,
+      "recon_trunc": 0.0,
+      "step": 400,
+      "student_loss": 0.04709221050143242,
+      "teacher_loss": 0.03479952737689018
+    },
+    {
+      "epoch": 0.11675675675675676,
+      "grad_norm": 3.46875,
+      "kd_loss": 0.010722252540290356,
+      "learning_rate": 8.571607968777349e-06,
+      "loss": 0.1873,
+      "recon_loss": 1.0106048583984375,
+      "recon_trunc": 0.0,
+      "step": 405,
+      "student_loss": 0.09067349135875702,
+      "teacher_loss": 0.08614400774240494
+    },
+    {
+      "epoch": 0.1181981981981982,
+      "grad_norm": 4.90625,
+      "kd_loss": 0.030308932065963745,
+      "learning_rate": 8.533624936070225e-06,
+      "loss": 0.1694,
+      "recon_loss": 1.0326484441757202,
+      "recon_trunc": 0.0,
+      "step": 410,
+      "student_loss": 0.07089435309171677,
+      "teacher_loss": 0.032449834048748016
+    },
+    {
+      "epoch": 0.11963963963963964,
+      "grad_norm": 3.0625,
+      "kd_loss": 0.01372783724218607,
+      "learning_rate": 8.495230061964289e-06,
+      "loss": 0.1837,
+      "recon_loss": 0.6862786412239075,
+      "recon_trunc": 0.0,
+      "step": 415,
+      "student_loss": 0.012874572537839413,
+      "teacher_loss": 0.012371852062642574
+    },
+    {
+      "epoch": 0.12108108108108108,
+      "grad_norm": 3.828125,
+      "kd_loss": 0.024159908294677734,
+      "learning_rate": 8.456427821354001e-06,
+      "loss": 0.1848,
+      "recon_loss": 0.2119012326002121,
+      "recon_trunc": 0.0,
+      "step": 420,
+      "student_loss": 0.012415669858455658,
+      "teacher_loss": 0.009588047862052917
+    },
+    {
+      "epoch": 0.12252252252252252,
+      "grad_norm": 2.875,
+      "kd_loss": 0.02482266165316105,
+      "learning_rate": 8.417222736612092e-06,
+      "loss": 0.1741,
+      "recon_loss": 1.655558466911316,
+      "recon_trunc": 0.0,
+      "step": 425,
+      "student_loss": 0.028823576867580414,
+      "teacher_loss": 0.03349754959344864
+    },
+    {
+      "epoch": 0.12396396396396396,
+      "grad_norm": 6.4375,
+      "kd_loss": 0.024806968867778778,
+      "learning_rate": 8.377619377062483e-06,
+      "loss": 0.1736,
+      "recon_loss": 0.437429279088974,
+      "recon_trunc": 0.0,
+      "step": 430,
+      "student_loss": 0.009496749378740788,
+      "teacher_loss": 0.00569483358412981
+    },
+    {
+      "epoch": 0.1254054054054054,
+      "grad_norm": 5.0625,
+      "kd_loss": 0.020807143300771713,
+      "learning_rate": 8.337622358447728e-06,
+      "loss": 0.1739,
+      "recon_loss": 1.1457756757736206,
+      "recon_trunc": 0.0,
+      "step": 435,
+      "student_loss": 0.061816953122615814,
+      "teacher_loss": 0.06000624969601631
+    },
+    {
+      "epoch": 0.12684684684684686,
+      "grad_norm": 3.328125,
+      "kd_loss": 0.0194116048514843,
+      "learning_rate": 8.29723634239106e-06,
+      "loss": 0.1738,
+      "recon_loss": 1.2350624799728394,
+      "recon_trunc": 0.0,
+      "step": 440,
+      "student_loss": 0.008348196744918823,
+      "teacher_loss": 0.004761967808008194
+    },
+    {
+      "epoch": 0.12828828828828828,
+      "grad_norm": 5.0,
+      "kd_loss": 0.02508135698735714,
+      "learning_rate": 8.256466035853077e-06,
+      "loss": 0.1692,
+      "recon_loss": 0.7592867016792297,
+      "recon_trunc": 0.0,
+      "step": 445,
+      "student_loss": 0.010359400883316994,
+      "teacher_loss": 0.005488038063049316
+    },
+    {
+      "epoch": 0.12972972972972974,
+      "grad_norm": 5.1875,
+      "kd_loss": 0.02380117028951645,
+      "learning_rate": 8.215316190583153e-06,
+      "loss": 0.1881,
+      "recon_loss": 0.5154184699058533,
+      "recon_trunc": 0.0,
+      "step": 450,
+      "student_loss": 0.018887203186750412,
+      "teacher_loss": 0.020634159445762634
+    },
+    {
+      "epoch": 0.13117117117117116,
+      "grad_norm": 4.40625,
+      "kd_loss": 0.0168263278901577,
+      "learning_rate": 8.173791602565622e-06,
+      "loss": 0.1951,
+      "recon_loss": 1.9520416259765625,
+      "recon_trunc": 0.0,
+      "step": 455,
+      "student_loss": 0.016898764297366142,
+      "teacher_loss": 0.011184322647750378
+    },
+    {
+      "epoch": 0.13261261261261262,
+      "grad_norm": 4.75,
+      "kd_loss": 0.013308191671967506,
+      "learning_rate": 8.13189711146081e-06,
+      "loss": 0.1663,
+      "recon_loss": 0.6237285733222961,
+      "recon_trunc": 0.0,
+      "step": 460,
+      "student_loss": 0.008111592382192612,
+      "teacher_loss": 0.005940328352153301
+    },
+    {
+      "epoch": 0.13405405405405404,
+      "grad_norm": 3.5,
+      "kd_loss": 0.03190253674983978,
+      "learning_rate": 8.08963760004098e-06,
+      "loss": 0.1856,
+      "recon_loss": 2.028031826019287,
+      "recon_trunc": 0.0,
+      "step": 465,
+      "student_loss": 0.05859745666384697,
+      "teacher_loss": 0.04754573106765747
+    },
+    {
+      "epoch": 0.1354954954954955,
+      "grad_norm": 3.21875,
+      "kd_loss": 0.019053883850574493,
+      "learning_rate": 8.047017993621242e-06,
+      "loss": 0.1901,
+      "recon_loss": 1.6077977418899536,
+      "recon_trunc": 0.0,
+      "step": 470,
+      "student_loss": 0.09916370362043381,
+      "teacher_loss": 0.1100371852517128
+    },
+    {
+      "epoch": 0.13693693693693693,
+      "grad_norm": 3.25,
+      "kd_loss": 0.022520730271935463,
+      "learning_rate": 8.004043259485519e-06,
+      "loss": 0.1813,
+      "recon_loss": 1.1354869604110718,
+      "recon_trunc": 0.0,
+      "step": 475,
+      "student_loss": 0.020928040146827698,
+      "teacher_loss": 0.017121572047472
+    },
+    {
+      "epoch": 0.13837837837837838,
+      "grad_norm": 3.953125,
+      "kd_loss": 0.03687773272395134,
+      "learning_rate": 7.960718406307602e-06,
+      "loss": 0.1814,
+      "recon_loss": 1.5030176639556885,
+      "recon_trunc": 0.1666666716337204,
+      "step": 480,
+      "student_loss": 0.04233800247311592,
+      "teacher_loss": 0.07700891047716141
+    },
+    {
+      "epoch": 0.1398198198198198,
+      "grad_norm": 3.6875,
+      "kd_loss": 0.02249104715883732,
+      "learning_rate": 7.917048483567408e-06,
+      "loss": 0.171,
+      "recon_loss": 2.312896490097046,
+      "recon_trunc": 0.0,
+      "step": 485,
+      "student_loss": 0.0480136014521122,
+      "teacher_loss": 0.011319181881844997
+    },
+    {
+      "epoch": 0.14126126126126126,
+      "grad_norm": 3.421875,
+      "kd_loss": 0.023666707798838615,
+      "learning_rate": 7.873038580962453e-06,
+      "loss": 0.1845,
+      "recon_loss": 1.3111594915390015,
+      "recon_trunc": 0.0,
+      "step": 490,
+      "student_loss": 0.02185920625925064,
+      "teacher_loss": 0.015433445572853088
+    },
+    {
+      "epoch": 0.14270270270270272,
+      "grad_norm": 2.703125,
+      "kd_loss": 0.025161005556583405,
+      "learning_rate": 7.828693827814662e-06,
+      "loss": 0.1668,
+      "recon_loss": 0.5953664183616638,
+      "recon_trunc": 0.0,
+      "step": 495,
+      "student_loss": 0.03152979165315628,
+      "teacher_loss": 0.009941949509084225
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 2.9375,
+      "kd_loss": 0.018994349986314774,
+      "learning_rate": 7.784019392472539e-06,
+      "loss": 0.1811,
+      "recon_loss": 1.3691961765289307,
+      "recon_trunc": 0.0,
+      "step": 500,
+      "student_loss": 0.010158221237361431,
+      "teacher_loss": 0.010299856774508953
+    },
+    {
+      "epoch": 0.1455855855855856,
+      "grad_norm": 2.96875,
+      "kd_loss": 0.022056058049201965,
+      "learning_rate": 7.739020481708816e-06,
+      "loss": 0.1726,
+      "recon_loss": 0.512145459651947,
+      "recon_trunc": 0.0,
+      "step": 505,
+      "student_loss": 0.04854091629385948,
+      "teacher_loss": 0.028620928525924683
+    },
+    {
+      "epoch": 0.14702702702702702,
+      "grad_norm": 3.328125,
+      "kd_loss": 0.020863963291049004,
+      "learning_rate": 7.693702340113587e-06,
+      "loss": 0.1572,
+      "recon_loss": 2.432304859161377,
+      "recon_trunc": 0.0,
+      "step": 510,
+      "student_loss": 0.011497420258820057,
+      "teacher_loss": 0.00617167167365551
+    },
+    {
+      "epoch": 0.14846846846846848,
+      "grad_norm": 2.4375,
+      "kd_loss": 0.020719008520245552,
+      "learning_rate": 7.648070249483074e-06,
+      "loss": 0.1642,
+      "recon_loss": 1.4956601858139038,
+      "recon_trunc": 0.0,
+      "step": 515,
+      "student_loss": 0.016647247597575188,
+      "teacher_loss": 0.012825959362089634
+    },
+    {
+      "epoch": 0.1499099099099099,
+      "grad_norm": 3.265625,
+      "kd_loss": 0.019960355013608932,
+      "learning_rate": 7.602129528204023e-06,
+      "loss": 0.1625,
+      "recon_loss": 1.01558518409729,
+      "recon_trunc": 0.0,
+      "step": 520,
+      "student_loss": 0.009609000757336617,
+      "teacher_loss": 0.004349292255938053
+    },
+    {
+      "epoch": 0.15135135135135136,
+      "grad_norm": 3.375,
+      "kd_loss": 0.034366805106401443,
+      "learning_rate": 7.555885530633864e-06,
+      "loss": 0.1601,
+      "recon_loss": 0.6185340285301208,
+      "recon_trunc": 0.0,
+      "step": 525,
+      "student_loss": 0.014354231767356396,
+      "teacher_loss": 0.007053965236991644
+    },
+    {
+      "epoch": 0.1527927927927928,
+      "grad_norm": 4.0625,
+      "kd_loss": 0.01583157293498516,
+      "learning_rate": 7.509343646476647e-06,
+      "loss": 0.1738,
+      "recon_loss": 1.1461244821548462,
+      "recon_trunc": 0.0,
+      "step": 530,
+      "student_loss": 0.14763206243515015,
+      "teacher_loss": 0.10371965914964676
+    },
+    {
+      "epoch": 0.15423423423423424,
+      "grad_norm": 3.1875,
+      "kd_loss": 0.023187261074781418,
+      "learning_rate": 7.462509300154892e-06,
+      "loss": 0.1697,
+      "recon_loss": 0.5664456486701965,
+      "recon_trunc": 0.0,
+      "step": 535,
+      "student_loss": 0.014488914050161839,
+      "teacher_loss": 0.03050735406577587
+    },
+    {
+      "epoch": 0.15567567567567567,
+      "grad_norm": 3.234375,
+      "kd_loss": 0.013575480319559574,
+      "learning_rate": 7.415387950177369e-06,
+      "loss": 0.151,
+      "recon_loss": 0.16455520689487457,
+      "recon_trunc": 0.0,
+      "step": 540,
+      "student_loss": 0.009347503073513508,
+      "teacher_loss": 0.0067581189796328545
+    },
+    {
+      "epoch": 0.15711711711711712,
+      "grad_norm": 2.140625,
+      "kd_loss": 0.027481766417622566,
+      "learning_rate": 7.367985088502914e-06,
+      "loss": 0.1635,
+      "recon_loss": 1.128605842590332,
+      "recon_trunc": 0.0,
+      "step": 545,
+      "student_loss": 0.04076429829001427,
+      "teacher_loss": 0.04589547961950302
+    },
+    {
+      "epoch": 0.15855855855855855,
+      "grad_norm": 2.5,
+      "kd_loss": 0.03289436176419258,
+      "learning_rate": 7.320306239900343e-06,
+      "loss": 0.1738,
+      "recon_loss": 0.9650799036026001,
+      "recon_trunc": 0.0,
+      "step": 550,
+      "student_loss": 0.03381229564547539,
+      "teacher_loss": 0.02062082104384899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.15625,
+      "kd_loss": 0.018133265897631645,
+      "learning_rate": 7.27235696130455e-06,
+      "loss": 0.1531,
+      "recon_loss": 0.411455899477005,
+      "recon_trunc": 0.0,
+      "step": 555,
+      "student_loss": 0.006184063386172056,
+      "teacher_loss": 0.0037621869705617428
+    },
+    {
+      "epoch": 0.16144144144144143,
+      "grad_norm": 5.6875,
+      "kd_loss": 0.02489805407822132,
+      "learning_rate": 7.2241428411688415e-06,
+      "loss": 0.172,
+      "recon_loss": 0.45451149344444275,
+      "recon_trunc": 0.0,
+      "step": 560,
+      "student_loss": 0.014761613681912422,
+      "teacher_loss": 0.012310503982007504
+    },
+    {
+      "epoch": 0.16288288288288288,
+      "grad_norm": 3.09375,
+      "kd_loss": 0.019155461341142654,
+      "learning_rate": 7.1756694988136165e-06,
+      "loss": 0.1716,
+      "recon_loss": 0.2166658192873001,
+      "recon_trunc": 0.0,
+      "step": 565,
+      "student_loss": 0.0065104472450912,
+      "teacher_loss": 0.008951494470238686
+    },
+    {
+      "epoch": 0.1643243243243243,
+      "grad_norm": 2.28125,
+      "kd_loss": 0.020588651299476624,
+      "learning_rate": 7.126942583771433e-06,
+      "loss": 0.1556,
+      "recon_loss": 0.6223914623260498,
+      "recon_trunc": 0.0,
+      "step": 570,
+      "student_loss": 0.0345882847905159,
+      "teacher_loss": 0.016995852813124657
+    },
+    {
+      "epoch": 0.16576576576576577,
+      "grad_norm": 5.28125,
+      "kd_loss": 0.014671197161078453,
+      "learning_rate": 7.077967775128558e-06,
+      "loss": 0.1652,
+      "recon_loss": 0.646995484828949,
+      "recon_trunc": 0.0,
+      "step": 575,
+      "student_loss": 0.00584026426076889,
+      "teacher_loss": 0.005363194737583399
+    },
+    {
+      "epoch": 0.16720720720720722,
+      "grad_norm": 4.3125,
+      "kd_loss": 0.025976594537496567,
+      "learning_rate": 7.028750780863078e-06,
+      "loss": 0.1515,
+      "recon_loss": 0.5720468163490295,
+      "recon_trunc": 0.0,
+      "step": 580,
+      "student_loss": 0.011792246252298355,
+      "teacher_loss": 0.012056869454681873
+    },
+    {
+      "epoch": 0.16864864864864865,
+      "grad_norm": 2.75,
+      "kd_loss": 0.030293436720967293,
+      "learning_rate": 6.979297337179632e-06,
+      "loss": 0.1529,
+      "recon_loss": 0.7066402435302734,
+      "recon_trunc": 0.0,
+      "step": 585,
+      "student_loss": 0.033711861819028854,
+      "teacher_loss": 0.011740412563085556
+    },
+    {
+      "epoch": 0.1700900900900901,
+      "grad_norm": 4.84375,
+      "kd_loss": 0.0336930938065052,
+      "learning_rate": 6.929613207840875e-06,
+      "loss": 0.157,
+      "recon_loss": 0.38057538866996765,
+      "recon_trunc": 0.13636364042758942,
+      "step": 590,
+      "student_loss": 0.011531775817275047,
+      "teacher_loss": 0.004667046945542097
+    },
+    {
+      "epoch": 0.17153153153153153,
+      "grad_norm": 4.15625,
+      "kd_loss": 0.01862204447388649,
+      "learning_rate": 6.8797041834956955e-06,
+      "loss": 0.1548,
+      "recon_loss": 0.25531551241874695,
+      "recon_trunc": 0.0,
+      "step": 595,
+      "student_loss": 0.005597068928182125,
+      "teacher_loss": 0.0031040518078953028
+    },
+    {
+      "epoch": 0.17297297297297298,
+      "grad_norm": 3.25,
+      "kd_loss": 0.022174149751663208,
+      "learning_rate": 6.829576081004339e-06,
+      "loss": 0.1515,
+      "recon_loss": 0.46167176961898804,
+      "recon_trunc": 0.0,
+      "step": 600,
+      "student_loss": 0.024388153105974197,
+      "teacher_loss": 0.023951301351189613
+    },
+    {
+      "epoch": 0.1744144144144144,
+      "grad_norm": 3.640625,
+      "kd_loss": 0.02048232965171337,
+      "learning_rate": 6.779234742760441e-06,
+      "loss": 0.1504,
+      "recon_loss": 0.217010498046875,
+      "recon_trunc": 0.0,
+      "step": 605,
+      "student_loss": 0.006499677896499634,
+      "teacher_loss": 0.004413343966007233
+    },
+    {
+      "epoch": 0.17585585585585586,
+      "grad_norm": 2.671875,
+      "kd_loss": 0.014861252158880234,
+      "learning_rate": 6.728686036010115e-06,
+      "loss": 0.1538,
+      "recon_loss": 0.5205051898956299,
+      "recon_trunc": 0.0,
+      "step": 610,
+      "student_loss": 0.004800914786756039,
+      "teacher_loss": 0.0033590400125831366
+    },
+    {
+      "epoch": 0.1772972972972973,
+      "grad_norm": 3.390625,
+      "kd_loss": 0.02478075958788395,
+      "learning_rate": 6.677935852168116e-06,
+      "loss": 0.1342,
+      "recon_loss": 0.4253961741924286,
+      "recon_trunc": 0.0,
+      "step": 615,
+      "student_loss": 0.015429160557687283,
+      "teacher_loss": 0.002787200966849923
+    },
+    {
+      "epoch": 0.17873873873873874,
+      "grad_norm": 3.28125,
+      "kd_loss": 0.030507735908031464,
+      "learning_rate": 6.62699010613121e-06,
+      "loss": 0.1527,
+      "recon_loss": 0.37841057777404785,
+      "recon_trunc": 0.0,
+      "step": 620,
+      "student_loss": 0.026107553392648697,
+      "teacher_loss": 0.015330830588936806
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 3.671875,
+      "kd_loss": 0.010809790343046188,
+      "learning_rate": 6.5758547355887944e-06,
+      "loss": 0.1368,
+      "recon_loss": 0.5961279273033142,
+      "recon_trunc": 0.0,
+      "step": 625,
+      "student_loss": 0.005336856469511986,
+      "teacher_loss": 0.005102858413010836
+    },
+    {
+      "epoch": 0.18162162162162163,
+      "grad_norm": 2.96875,
+      "kd_loss": 0.020181534811854362,
+      "learning_rate": 6.524535700330858e-06,
+      "loss": 0.1416,
+      "recon_loss": 0.7704125046730042,
+      "recon_trunc": 0.0,
+      "step": 630,
+      "student_loss": 0.06882191449403763,
+      "teacher_loss": 0.06232360750436783
+    },
+    {
+      "epoch": 0.18306306306306305,
+      "grad_norm": 2.3125,
+      "kd_loss": 0.02182999812066555,
+      "learning_rate": 6.4730389815533865e-06,
+      "loss": 0.131,
+      "recon_loss": 0.7982496023178101,
+      "recon_trunc": 0.0,
+      "step": 635,
+      "student_loss": 0.013796217739582062,
+      "teacher_loss": 0.008559397421777248
+    },
+    {
+      "epoch": 0.1845045045045045,
+      "grad_norm": 2.953125,
+      "kd_loss": 0.01843719743192196,
+      "learning_rate": 6.421370581161244e-06,
+      "loss": 0.1366,
+      "recon_loss": 0.6617377400398254,
+      "recon_trunc": 0.0,
+      "step": 640,
+      "student_loss": 0.02629081718623638,
+      "teacher_loss": 0.01877058483660221
+    },
+    {
+      "epoch": 0.18594594594594593,
+      "grad_norm": 2.84375,
+      "kd_loss": 0.020510626956820488,
+      "learning_rate": 6.36953652106866e-06,
+      "loss": 0.159,
+      "recon_loss": 0.4313066005706787,
+      "recon_trunc": 0.0,
+      "step": 645,
+      "student_loss": 0.014318160712718964,
+      "teacher_loss": 0.010488627478480339
+    },
+    {
+      "epoch": 0.1873873873873874,
+      "grad_norm": 2.65625,
+      "kd_loss": 0.020855048671364784,
+      "learning_rate": 6.317542842497384e-06,
+      "loss": 0.1485,
+      "recon_loss": 0.4753470718860626,
+      "recon_trunc": 0.0,
+      "step": 650,
+      "student_loss": 0.012120629660785198,
+      "teacher_loss": 0.01488297339528799
+    },
+    {
+      "epoch": 0.18882882882882882,
+      "grad_norm": 2.5,
+      "kd_loss": 0.022243686020374298,
+      "learning_rate": 6.265395605272581e-06,
+      "loss": 0.1363,
+      "recon_loss": 0.2491840422153473,
+      "recon_trunc": 0.0,
+      "step": 655,
+      "student_loss": 0.020461823791265488,
+      "teacher_loss": 0.012900102883577347
+    },
+    {
+      "epoch": 0.19027027027027027,
+      "grad_norm": 2.65625,
+      "kd_loss": 0.017112459987401962,
+      "learning_rate": 6.213100887116568e-06,
+      "loss": 0.1552,
+      "recon_loss": 0.4894320070743561,
+      "recon_trunc": 0.0,
+      "step": 660,
+      "student_loss": 0.02942224219441414,
+      "teacher_loss": 0.022297410294413567
+    },
+    {
+      "epoch": 0.19171171171171172,
+      "grad_norm": 2.984375,
+      "kd_loss": 0.01707088015973568,
+      "learning_rate": 6.1606647829404485e-06,
+      "loss": 0.1553,
+      "recon_loss": 0.13572968542575836,
+      "recon_trunc": 0.0,
+      "step": 665,
+      "student_loss": 0.012287209741771221,
+      "teacher_loss": 0.005970606114715338
+    },
+    {
+      "epoch": 0.19315315315315315,
+      "grad_norm": 3.40625,
+      "kd_loss": 0.020078979432582855,
+      "learning_rate": 6.108093404133772e-06,
+      "loss": 0.158,
+      "recon_loss": 0.9134924411773682,
+      "recon_trunc": 0.0,
+      "step": 670,
+      "student_loss": 0.006748141720890999,
+      "teacher_loss": 0.007465692702680826
+    },
+    {
+      "epoch": 0.1945945945945946,
+      "grad_norm": 2.03125,
+      "kd_loss": 0.02451922371983528,
+      "learning_rate": 6.0553928778522405e-06,
+      "loss": 0.1531,
+      "recon_loss": 0.7074742913246155,
+      "recon_trunc": 0.0,
+      "step": 675,
+      "student_loss": 0.02256634272634983,
+      "teacher_loss": 0.020620204508304596
+    },
+    {
+      "epoch": 0.19603603603603603,
+      "grad_norm": 6.625,
+      "kd_loss": 0.01408644299954176,
+      "learning_rate": 6.002569346303605e-06,
+      "loss": 0.1731,
+      "recon_loss": 0.37631916999816895,
+      "recon_trunc": 0.0,
+      "step": 680,
+      "student_loss": 0.0062326169572770596,
+      "teacher_loss": 0.017630083486437798
+    },
+    {
+      "epoch": 0.1974774774774775,
+      "grad_norm": 4.71875,
+      "kd_loss": 0.02016598917543888,
+      "learning_rate": 5.949628966031785e-06,
+      "loss": 0.1428,
+      "recon_loss": 0.4535362422466278,
+      "recon_trunc": 0.0,
+      "step": 685,
+      "student_loss": 0.008984028361737728,
+      "teacher_loss": 0.004230538848787546
+    },
+    {
+      "epoch": 0.1989189189189189,
+      "grad_norm": 3.765625,
+      "kd_loss": 0.012536494061350822,
+      "learning_rate": 5.8965779071993365e-06,
+      "loss": 0.149,
+      "recon_loss": 0.8645880818367004,
+      "recon_trunc": 0.0,
+      "step": 690,
+      "student_loss": 0.05488922819495201,
+      "teacher_loss": 0.03522896021604538
+    },
+    {
+      "epoch": 0.20036036036036037,
+      "grad_norm": 2.078125,
+      "kd_loss": 0.010920881293714046,
+      "learning_rate": 5.843422352868321e-06,
+      "loss": 0.1423,
+      "recon_loss": 2.1881675720214844,
+      "recon_trunc": 0.0,
+      "step": 695,
+      "student_loss": 0.004224100150167942,
+      "teacher_loss": 0.0032876133918762207
+    },
+    {
+      "epoch": 0.2018018018018018,
+      "grad_norm": 2.53125,
+      "kd_loss": 0.013865644112229347,
+      "learning_rate": 5.7901684982796716e-06,
+      "loss": 0.146,
+      "recon_loss": 0.7409210801124573,
+      "recon_trunc": 0.0,
+      "step": 700,
+      "student_loss": 0.06459976732730865,
+      "teacher_loss": 0.07415220141410828
+    },
+    {
+      "epoch": 0.20324324324324325,
+      "grad_norm": 2.5625,
+      "kd_loss": 0.01686241291463375,
+      "learning_rate": 5.736822550131147e-06,
+      "loss": 0.146,
+      "recon_loss": 0.6043723821640015,
+      "recon_trunc": 0.0,
+      "step": 705,
+      "student_loss": 0.0394061915576458,
+      "teacher_loss": 0.046153660863637924
+    },
+    {
+      "epoch": 0.20468468468468468,
+      "grad_norm": 2.75,
+      "kd_loss": 0.01901497319340706,
+      "learning_rate": 5.68339072585394e-06,
+      "loss": 0.1342,
+      "recon_loss": 0.5759344696998596,
+      "recon_trunc": 0.0,
+      "step": 710,
+      "student_loss": 0.05447360873222351,
+      "teacher_loss": 0.047757674008607864
+    },
+    {
+      "epoch": 0.20612612612612613,
+      "grad_norm": 2.8125,
+      "kd_loss": 0.011764246970415115,
+      "learning_rate": 5.629879252888046e-06,
+      "loss": 0.1455,
+      "recon_loss": 0.1448301523923874,
+      "recon_trunc": 0.0,
+      "step": 715,
+      "student_loss": 0.005854414775967598,
+      "teacher_loss": 0.0046508414670825005
+    },
+    {
+      "epoch": 0.20756756756756756,
+      "grad_norm": 2.578125,
+      "kd_loss": 0.014321034774184227,
+      "learning_rate": 5.576294367956455e-06,
+      "loss": 0.1433,
+      "recon_loss": 0.3958251476287842,
+      "recon_trunc": 0.0,
+      "step": 720,
+      "student_loss": 0.00420423923060298,
+      "teacher_loss": 0.0038505124393850565
+    },
+    {
+      "epoch": 0.209009009009009,
+      "grad_norm": 2.859375,
+      "kd_loss": 0.009370187297463417,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.1544,
+      "recon_loss": 0.023905467242002487,
+      "recon_trunc": 0.0,
+      "step": 725,
+      "student_loss": 0.009239193983376026,
+      "teacher_loss": 0.007000849116593599
+    },
+    {
+      "epoch": 0.21045045045045044,
+      "grad_norm": 4.5,
+      "kd_loss": 0.014552440494298935,
+      "learning_rate": 5.4689293511408155e-06,
+      "loss": 0.1503,
+      "recon_loss": 0.366656094789505,
+      "recon_trunc": 0.0,
+      "step": 730,
+      "student_loss": 0.005608205683529377,
+      "teacher_loss": 0.004755438771098852
+    },
+    {
+      "epoch": 0.2118918918918919,
+      "grad_norm": 2.59375,
+      "kd_loss": 0.017068471759557724,
+      "learning_rate": 5.4151617325708595e-06,
+      "loss": 0.1447,
+      "recon_loss": 0.7658705115318298,
+      "recon_trunc": 0.0,
+      "step": 735,
+      "student_loss": 0.04276591166853905,
+      "teacher_loss": 0.030612222850322723
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 2.65625,
+      "kd_loss": 0.025976940989494324,
+      "learning_rate": 5.361345727204972e-06,
+      "loss": 0.1501,
+      "recon_loss": 0.8174273371696472,
+      "recon_trunc": 0.02857142873108387,
+      "step": 740,
+      "student_loss": 0.06527666747570038,
+      "teacher_loss": 0.041367143392562866
+    },
+    {
+      "epoch": 0.21477477477477477,
+      "grad_norm": 3.375,
+      "kd_loss": 0.01059594377875328,
+      "learning_rate": 5.307487607259175e-06,
+      "loss": 0.1492,
+      "recon_loss": 0.48267030715942383,
+      "recon_trunc": 0.0,
+      "step": 745,
+      "student_loss": 0.07378892600536346,
+      "teacher_loss": 0.028776884078979492
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 2.578125,
+      "kd_loss": 0.013439241796731949,
+      "learning_rate": 5.253593649857907e-06,
+      "loss": 0.1426,
+      "recon_loss": 0.6058846116065979,
+      "recon_trunc": 0.0,
+      "step": 750,
+      "student_loss": 0.004285452887415886,
+      "teacher_loss": 0.0033083639573305845
+    },
+    {
+      "epoch": 0.21765765765765765,
+      "grad_norm": 2.25,
+      "kd_loss": 0.03229266405105591,
+      "learning_rate": 5.199670136302443e-06,
+      "loss": 0.1494,
+      "recon_loss": 0.7599271535873413,
+      "recon_trunc": 0.0,
+      "step": 755,
+      "student_loss": 0.03914610669016838,
+      "teacher_loss": 0.02319481037557125
+    },
+    {
+      "epoch": 0.2190990990990991,
+      "grad_norm": 2.828125,
+      "kd_loss": 0.029138775542378426,
+      "learning_rate": 5.145723351338799e-06,
+      "loss": 0.141,
+      "recon_loss": 0.4348209500312805,
+      "recon_trunc": 0.0,
+      "step": 760,
+      "student_loss": 0.006047646515071392,
+      "teacher_loss": 0.0045949919149279594
+    },
+    {
+      "epoch": 0.22054054054054054,
+      "grad_norm": 2.96875,
+      "kd_loss": 0.014339631423354149,
+      "learning_rate": 5.091759582425263e-06,
+      "loss": 0.1622,
+      "recon_loss": 0.8768014907836914,
+      "recon_trunc": 0.0,
+      "step": 765,
+      "student_loss": 0.004327854607254267,
+      "teacher_loss": 0.0032785937655717134
+    },
+    {
+      "epoch": 0.221981981981982,
+      "grad_norm": 2.734375,
+      "kd_loss": 0.017597543075680733,
+      "learning_rate": 5.037785118999585e-06,
+      "loss": 0.1527,
+      "recon_loss": 1.3829718828201294,
+      "recon_trunc": 0.0,
+      "step": 770,
+      "student_loss": 0.021684447303414345,
+      "teacher_loss": 0.013488702476024628
+    },
+    {
+      "epoch": 0.22342342342342342,
+      "grad_norm": 3.125,
+      "kd_loss": 0.011236852034926414,
+      "learning_rate": 4.983806251745958e-06,
+      "loss": 0.16,
+      "recon_loss": 1.3483177423477173,
+      "recon_trunc": 0.0,
+      "step": 775,
+      "student_loss": 0.005938877817243338,
+      "teacher_loss": 0.013820645399391651
+    },
+    {
+      "epoch": 0.22486486486486487,
+      "grad_norm": 5.8125,
+      "kd_loss": 0.013352782465517521,
+      "learning_rate": 4.9298292718618295e-06,
+      "loss": 0.1497,
+      "recon_loss": 0.6016652584075928,
+      "recon_trunc": 0.0,
+      "step": 780,
+      "student_loss": 0.012808529660105705,
+      "teacher_loss": 0.006906051654368639
+    },
+    {
+      "epoch": 0.2263063063063063,
+      "grad_norm": 3.0,
+      "kd_loss": 0.01723395846784115,
+      "learning_rate": 4.875860470324685e-06,
+      "loss": 0.1528,
+      "recon_loss": 0.37664490938186646,
+      "recon_trunc": 0.0,
+      "step": 785,
+      "student_loss": 0.0065705315209925175,
+      "teacher_loss": 0.006534716580063105
+    },
+    {
+      "epoch": 0.22774774774774775,
+      "grad_norm": 3.5,
+      "kd_loss": 0.017751790583133698,
+      "learning_rate": 4.821906137158822e-06,
+      "loss": 0.1391,
+      "recon_loss": 0.8859590291976929,
+      "recon_trunc": 0.0,
+      "step": 790,
+      "student_loss": 0.012177354656159878,
+      "teacher_loss": 0.011239876039326191
+    },
+    {
+      "epoch": 0.22918918918918918,
+      "grad_norm": 6.5625,
+      "kd_loss": 0.017111526802182198,
+      "learning_rate": 4.767972560702266e-06,
+      "loss": 0.1482,
+      "recon_loss": 1.6100565195083618,
+      "recon_trunc": 0.0,
+      "step": 795,
+      "student_loss": 0.14045321941375732,
+      "teacher_loss": 0.1274717152118683
+    },
+    {
+      "epoch": 0.23063063063063063,
+      "grad_norm": 3.0,
+      "kd_loss": 0.02000945806503296,
+      "learning_rate": 4.7140660268738565e-06,
+      "loss": 0.141,
+      "recon_loss": 0.6955020427703857,
+      "recon_trunc": 0.0,
+      "step": 800,
+      "student_loss": 0.0039167930372059345,
+      "teacher_loss": 0.018537431955337524
+    },
+    {
+      "epoch": 0.23207207207207206,
+      "grad_norm": 2.46875,
+      "kd_loss": 0.013010330498218536,
+      "learning_rate": 4.660192818440642e-06,
+      "loss": 0.1426,
+      "recon_loss": 0.6611738801002502,
+      "recon_trunc": 0.0,
+      "step": 805,
+      "student_loss": 0.00421526376157999,
+      "teacher_loss": 0.005498496349900961
+    },
+    {
+      "epoch": 0.23351351351351352,
+      "grad_norm": 2.546875,
+      "kd_loss": 0.019325681030750275,
+      "learning_rate": 4.606359214285615e-06,
+      "loss": 0.1464,
+      "recon_loss": 0.44592851400375366,
+      "recon_trunc": 0.0,
+      "step": 810,
+      "student_loss": 0.004032685421407223,
+      "teacher_loss": 0.0021672367583960295
+    },
+    {
+      "epoch": 0.23495495495495494,
+      "grad_norm": 3.40625,
+      "kd_loss": 0.015629149973392487,
+      "learning_rate": 4.5525714886759214e-06,
+      "loss": 0.1334,
+      "recon_loss": 2.691049098968506,
+      "recon_trunc": 0.0,
+      "step": 815,
+      "student_loss": 0.004834800027310848,
+      "teacher_loss": 0.005103384144604206
+    },
+    {
+      "epoch": 0.2363963963963964,
+      "grad_norm": 2.828125,
+      "kd_loss": 0.008663350716233253,
+      "learning_rate": 4.498835910531595e-06,
+      "loss": 0.1547,
+      "recon_loss": 2.215418577194214,
+      "recon_trunc": 0.0,
+      "step": 820,
+      "student_loss": 0.005945620592683554,
+      "teacher_loss": 0.005259233992546797
+    },
+    {
+      "epoch": 0.23783783783783785,
+      "grad_norm": 2.421875,
+      "kd_loss": 0.017225855961441994,
+      "learning_rate": 4.445158742694919e-06,
+      "loss": 0.1415,
+      "recon_loss": 0.08885490894317627,
+      "recon_trunc": 0.0,
+      "step": 825,
+      "student_loss": 0.008877439424395561,
+      "teacher_loss": 0.00699736550450325
+    },
+    {
+      "epoch": 0.23927927927927928,
+      "grad_norm": 2.921875,
+      "kd_loss": 0.02307087369263172,
+      "learning_rate": 4.391546241200503e-06,
+      "loss": 0.1143,
+      "recon_loss": 0.5780781507492065,
+      "recon_trunc": 0.0,
+      "step": 830,
+      "student_loss": 0.019044674932956696,
+      "teacher_loss": 0.010328766889870167
+    },
+    {
+      "epoch": 0.24072072072072073,
+      "grad_norm": 1.984375,
+      "kd_loss": 0.019833428785204887,
+      "learning_rate": 4.338004654546136e-06,
+      "loss": 0.1375,
+      "recon_loss": 0.7510714530944824,
+      "recon_trunc": 0.095238097012043,
+      "step": 835,
+      "student_loss": 0.07427078485488892,
+      "teacher_loss": 0.07200340926647186
+    },
+    {
+      "epoch": 0.24216216216216216,
+      "grad_norm": 2.234375,
+      "kd_loss": 0.015543047338724136,
+      "learning_rate": 4.2845402229645366e-06,
+      "loss": 0.1257,
+      "recon_loss": 1.0261766910552979,
+      "recon_trunc": 0.0,
+      "step": 840,
+      "student_loss": 0.006065291352570057,
+      "teacher_loss": 0.03649454191327095
+    },
+    {
+      "epoch": 0.2436036036036036,
+      "grad_norm": 2.78125,
+      "kd_loss": 0.02092084288597107,
+      "learning_rate": 4.231159177696057e-06,
+      "loss": 0.1441,
+      "recon_loss": 1.2153651714324951,
+      "recon_trunc": 0.0,
+      "step": 845,
+      "student_loss": 0.063958078622818,
+      "teacher_loss": 0.034890804439783096
+    },
+    {
+      "epoch": 0.24504504504504504,
+      "grad_norm": 3.09375,
+      "kd_loss": 0.010684914886951447,
+      "learning_rate": 4.177867740262437e-06,
+      "loss": 0.1424,
+      "recon_loss": 0.30497077107429504,
+      "recon_trunc": 0.0,
+      "step": 850,
+      "student_loss": 0.0678933784365654,
+      "teacher_loss": 0.04117945581674576
+    },
+    {
+      "epoch": 0.2464864864864865,
+      "grad_norm": 1.9296875,
+      "kd_loss": 0.012471984140574932,
+      "learning_rate": 4.1246721217416846e-06,
+      "loss": 0.1421,
+      "recon_loss": 1.1690119504928589,
+      "recon_trunc": 0.0,
+      "step": 855,
+      "student_loss": 0.03430585563182831,
+      "teacher_loss": 0.01472096610814333
+    },
+    {
+      "epoch": 0.24792792792792792,
+      "grad_norm": 3.0625,
+      "kd_loss": 0.018260203301906586,
+      "learning_rate": 4.071578522044188e-06,
+      "loss": 0.1476,
+      "recon_loss": 0.43741315603256226,
+      "recon_trunc": 0.0,
+      "step": 860,
+      "student_loss": 0.011916948482394218,
+      "teacher_loss": 0.019032137468457222
+    },
+    {
+      "epoch": 0.24936936936936938,
+      "grad_norm": 3.140625,
+      "kd_loss": 0.018417567014694214,
+      "learning_rate": 4.018593129190113e-06,
+      "loss": 0.1604,
+      "recon_loss": 0.4862760901451111,
+      "recon_trunc": 0.0,
+      "step": 865,
+      "student_loss": 0.010221644304692745,
+      "teacher_loss": 0.006534127052873373
+    },
+    {
+      "epoch": 0.2508108108108108,
+      "grad_norm": 2.3125,
+      "kd_loss": 0.008507420308887959,
+      "learning_rate": 3.965722118588199e-06,
+      "loss": 0.1298,
+      "recon_loss": 0.001144498586654663,
+      "recon_trunc": 0.0,
+      "step": 870,
+      "student_loss": 0.004630399402230978,
+      "teacher_loss": 0.0041196770034730434
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 3.453125,
+      "kd_loss": 0.01141602173447609,
+      "learning_rate": 3.912971652316017e-06,
+      "loss": 0.1448,
+      "recon_loss": 0.3633114695549011,
+      "recon_trunc": 0.0,
+      "step": 875,
+      "student_loss": 0.004442453384399414,
+      "teacher_loss": 0.0030530148651450872
+    },
+    {
+      "epoch": 0.2536936936936937,
+      "grad_norm": 2.3125,
+      "kd_loss": 0.036327384412288666,
+      "learning_rate": 3.8603478784017845e-06,
+      "loss": 0.1523,
+      "recon_loss": 1.515076756477356,
+      "recon_trunc": 0.0,
+      "step": 880,
+      "student_loss": 0.04675779491662979,
+      "teacher_loss": 0.0254712775349617
+    },
+    {
+      "epoch": 0.25513513513513514,
+      "grad_norm": 2.75,
+      "kd_loss": 0.008054370060563087,
+      "learning_rate": 3.8078569301078165e-06,
+      "loss": 0.1396,
+      "recon_loss": 0.585823118686676,
+      "recon_trunc": 0.0,
+      "step": 885,
+      "student_loss": 0.054446298629045486,
+      "teacher_loss": 0.04862826317548752
+    },
+    {
+      "epoch": 0.25657657657657656,
+      "grad_norm": 2.453125,
+      "kd_loss": 0.010570000857114792,
+      "learning_rate": 3.7555049252157006e-06,
+      "loss": 0.1467,
+      "recon_loss": 0.48317578434944153,
+      "recon_trunc": 0.0,
+      "step": 890,
+      "student_loss": 0.003888532752171159,
+      "teacher_loss": 0.004439337644726038
+    },
+    {
+      "epoch": 0.258018018018018,
+      "grad_norm": 2.578125,
+      "kd_loss": 0.01936684176325798,
+      "learning_rate": 3.703297965313275e-06,
+      "loss": 0.1415,
+      "recon_loss": 0.8492840528488159,
+      "recon_trunc": 0.0,
+      "step": 895,
+      "student_loss": 0.059838131070137024,
+      "teacher_loss": 0.039372529834508896
+    },
+    {
+      "epoch": 0.2594594594594595,
+      "grad_norm": 2.578125,
+      "kd_loss": 0.01827370561659336,
+      "learning_rate": 3.651242135083485e-06,
+      "loss": 0.1256,
+      "recon_loss": 0.4612889885902405,
+      "recon_trunc": 0.0,
+      "step": 900,
+      "student_loss": 0.014989878982305527,
+      "teacher_loss": 0.02137553133070469
+    },
+    {
+      "epoch": 0.2609009009009009,
+      "grad_norm": 2.234375,
+      "kd_loss": 0.012334790080785751,
+      "learning_rate": 3.599343501595229e-06,
+      "loss": 0.1431,
+      "recon_loss": 1.2446216344833374,
+      "recon_trunc": 0.0,
+      "step": 905,
+      "student_loss": 0.00881471298635006,
+      "teacher_loss": 0.009096480906009674
+    },
+    {
+      "epoch": 0.2623423423423423,
+      "grad_norm": 2.921875,
+      "kd_loss": 0.024297041818499565,
+      "learning_rate": 3.5476081135962335e-06,
+      "loss": 0.1526,
+      "recon_loss": 1.1183922290802002,
+      "recon_trunc": 0.0,
+      "step": 910,
+      "student_loss": 0.17456559836864471,
+      "teacher_loss": 0.11914648860692978
+    },
+    {
+      "epoch": 0.2637837837837838,
+      "grad_norm": 3.171875,
+      "kd_loss": 0.010026839561760426,
+      "learning_rate": 3.496042000808091e-06,
+      "loss": 0.1452,
+      "recon_loss": 1.7595101594924927,
+      "recon_trunc": 0.0,
+      "step": 915,
+      "student_loss": 0.004593400750309229,
+      "teacher_loss": 0.007123799063265324
+    },
+    {
+      "epoch": 0.26522522522522524,
+      "grad_norm": 3.140625,
+      "kd_loss": 0.008071749471127987,
+      "learning_rate": 3.4446511732234846e-06,
+      "loss": 0.1456,
+      "recon_loss": 2.1998934745788574,
+      "recon_trunc": 0.0,
+      "step": 920,
+      "student_loss": 0.004290066193789244,
+      "teacher_loss": 0.003803086467087269
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 2.96875,
+      "kd_loss": 0.017971698194742203,
+      "learning_rate": 3.3934416204057396e-06,
+      "loss": 0.1329,
+      "recon_loss": 0.7439331412315369,
+      "recon_trunc": 0.0,
+      "step": 925,
+      "student_loss": 0.04140475019812584,
+      "teacher_loss": 0.0262799970805645
+    },
+    {
+      "epoch": 0.2681081081081081,
+      "grad_norm": 2.796875,
+      "kd_loss": 0.010629110969603062,
+      "learning_rate": 3.3424193107907386e-06,
+      "loss": 0.1406,
+      "recon_loss": 0.2620896100997925,
+      "recon_trunc": 0.0,
+      "step": 930,
+      "student_loss": 0.003074842505156994,
+      "teacher_loss": 0.002547505311667919
+    },
+    {
+      "epoch": 0.26954954954954957,
+      "grad_norm": 2.953125,
+      "kd_loss": 0.017617138102650642,
+      "learning_rate": 3.291590190991306e-06,
+      "loss": 0.1305,
+      "recon_loss": 0.4242994487285614,
+      "recon_trunc": 0.0,
+      "step": 935,
+      "student_loss": 0.008209005929529667,
+      "teacher_loss": 0.0033676454331725836
+    },
+    {
+      "epoch": 0.270990990990991,
+      "grad_norm": 2.15625,
+      "kd_loss": 0.01549585908651352,
+      "learning_rate": 3.240960185104137e-06,
+      "loss": 0.1375,
+      "recon_loss": 1.0953670740127563,
+      "recon_trunc": 0.0,
+      "step": 940,
+      "student_loss": 0.03813717141747475,
+      "teacher_loss": 0.012444056570529938
+    },
+    {
+      "epoch": 0.2724324324324324,
+      "grad_norm": 2.265625,
+      "kd_loss": 0.010429909452795982,
+      "learning_rate": 3.1905351940193417e-06,
+      "loss": 0.1433,
+      "recon_loss": 1.7625950574874878,
+      "recon_trunc": 0.0,
+      "step": 945,
+      "student_loss": 0.0326802097260952,
+      "teacher_loss": 0.026324408128857613
+    },
+    {
+      "epoch": 0.27387387387387385,
+      "grad_norm": 3.078125,
+      "kd_loss": 0.007424323353916407,
+      "learning_rate": 3.14032109473271e-06,
+      "loss": 0.1658,
+      "recon_loss": 2.978121519088745,
+      "recon_trunc": 0.0,
+      "step": 950,
+      "student_loss": 0.010005434043705463,
+      "teacher_loss": 0.008725137449800968
+    },
+    {
+      "epoch": 0.27531531531531533,
+      "grad_norm": 3.71875,
+      "kd_loss": 0.013719601556658745,
+      "learning_rate": 3.090323739660742e-06,
+      "loss": 0.1171,
+      "recon_loss": 0.39180198311805725,
+      "recon_trunc": 0.0,
+      "step": 955,
+      "student_loss": 0.00335954362526536,
+      "teacher_loss": 0.0019092520233243704
+    },
+    {
+      "epoch": 0.27675675675675676,
+      "grad_norm": 2.671875,
+      "kd_loss": 0.012734523974359035,
+      "learning_rate": 3.040548955958563e-06,
+      "loss": 0.139,
+      "recon_loss": 0.5425288081169128,
+      "recon_trunc": 0.0,
+      "step": 960,
+      "student_loss": 0.04997498542070389,
+      "teacher_loss": 0.01856311596930027
+    },
+    {
+      "epoch": 0.2781981981981982,
+      "grad_norm": 2.890625,
+      "kd_loss": 0.007673286367207766,
+      "learning_rate": 2.991002544840763e-06,
+      "loss": 0.1363,
+      "recon_loss": 3.391064167022705,
+      "recon_trunc": 0.0,
+      "step": 965,
+      "student_loss": 0.010929310694336891,
+      "teacher_loss": 0.010569097474217415
+    },
+    {
+      "epoch": 0.2796396396396396,
+      "grad_norm": 2.3125,
+      "kd_loss": 0.014591219834983349,
+      "learning_rate": 2.9416902809052817e-06,
+      "loss": 0.1324,
+      "recon_loss": 0.7789583206176758,
+      "recon_trunc": 0.0,
+      "step": 970,
+      "student_loss": 0.0037616032641381025,
+      "teacher_loss": 0.002346796914935112
+    },
+    {
+      "epoch": 0.2810810810810811,
+      "grad_norm": 2.984375,
+      "kd_loss": 0.018719913437962532,
+      "learning_rate": 2.8926179114603704e-06,
+      "loss": 0.1492,
+      "recon_loss": 0.3576004207134247,
+      "recon_trunc": 0.0,
+      "step": 975,
+      "student_loss": 0.0052808732725679874,
+      "teacher_loss": 0.0016440697945654392
+    },
+    {
+      "epoch": 0.2825225225225225,
+      "grad_norm": 2.46875,
+      "kd_loss": 0.02334708347916603,
+      "learning_rate": 2.8437911558547515e-06,
+      "loss": 0.1286,
+      "recon_loss": 1.4327666759490967,
+      "recon_trunc": 0.0,
+      "step": 980,
+      "student_loss": 0.03398752585053444,
+      "teacher_loss": 0.02993309311568737
+    },
+    {
+      "epoch": 0.28396396396396395,
+      "grad_norm": 2.625,
+      "kd_loss": 0.01443286333233118,
+      "learning_rate": 2.7952157048110406e-06,
+      "loss": 0.1337,
+      "recon_loss": 1.386612057685852,
+      "recon_trunc": 0.0,
+      "step": 985,
+      "student_loss": 0.0026713418774306774,
+      "teacher_loss": 0.002368104876950383
+    },
+    {
+      "epoch": 0.28540540540540543,
+      "grad_norm": 3.390625,
+      "kd_loss": 0.0160578191280365,
+      "learning_rate": 2.74689721976249e-06,
+      "loss": 0.1349,
+      "recon_loss": 0.6237896680831909,
+      "recon_trunc": 0.0,
+      "step": 990,
+      "student_loss": 0.008113768883049488,
+      "teacher_loss": 0.011220441199839115
+    },
+    {
+      "epoch": 0.28684684684684686,
+      "grad_norm": 3.203125,
+      "kd_loss": 0.020893100649118423,
+      "learning_rate": 2.698841332193148e-06,
+      "loss": 0.1331,
+      "recon_loss": 0.15756909549236298,
+      "recon_trunc": 0.0,
+      "step": 995,
+      "student_loss": 0.06304652988910675,
+      "teacher_loss": 0.02618415839970112
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 1.9765625,
+      "kd_loss": 0.02194441296160221,
+      "learning_rate": 2.6510536429815224e-06,
+      "loss": 0.1284,
+      "recon_loss": 0.6216073632240295,
+      "recon_trunc": 0.0,
+      "step": 1000,
+      "student_loss": 0.01729448512196541,
+      "teacher_loss": 0.011490650475025177
+    },
+    {
+      "epoch": 0.2897297297297297,
+      "grad_norm": 2.640625,
+      "kd_loss": 0.013767369091510773,
+      "learning_rate": 2.6035397217477977e-06,
+      "loss": 0.1436,
+      "recon_loss": 0.8209414482116699,
+      "recon_trunc": 0.0,
+      "step": 1005,
+      "student_loss": 0.03666730597615242,
+      "teacher_loss": 0.01536045502871275
+    },
+    {
+      "epoch": 0.2911711711711712,
+      "grad_norm": 3.09375,
+      "kd_loss": 0.011710915714502335,
+      "learning_rate": 2.5563051062047034e-06,
+      "loss": 0.1338,
+      "recon_loss": 2.728196382522583,
+      "recon_trunc": 0.0,
+      "step": 1010,
+      "student_loss": 0.0037028242368251085,
+      "teacher_loss": 0.0026822902727872133
+    },
+    {
+      "epoch": 0.2926126126126126,
+      "grad_norm": 3.03125,
+      "kd_loss": 0.011254915967583656,
+      "learning_rate": 2.5093553015120937e-06,
+      "loss": 0.1337,
+      "recon_loss": 0.05140983313322067,
+      "recon_trunc": 0.0,
+      "step": 1015,
+      "student_loss": 0.005124497693032026,
+      "teacher_loss": 0.003702581627294421
+    },
+    {
+      "epoch": 0.29405405405405405,
+      "grad_norm": 2.265625,
+      "kd_loss": 0.019891448318958282,
+      "learning_rate": 2.4626957796353257e-06,
+      "loss": 0.1446,
+      "recon_loss": 0.6191113591194153,
+      "recon_trunc": 0.0,
+      "step": 1020,
+      "student_loss": 0.04158704727888107,
+      "teacher_loss": 0.06546971201896667
+    },
+    {
+      "epoch": 0.2954954954954955,
+      "grad_norm": 2.515625,
+      "kd_loss": 0.010566615499556065,
+      "learning_rate": 2.416331978707506e-06,
+      "loss": 0.1418,
+      "recon_loss": 1.0828590393066406,
+      "recon_trunc": 0.0,
+      "step": 1025,
+      "student_loss": 0.04541200399398804,
+      "teacher_loss": 0.03305817022919655
+    },
+    {
+      "epoch": 0.29693693693693696,
+      "grad_norm": 2.046875,
+      "kd_loss": 0.02790280058979988,
+      "learning_rate": 2.3702693023956853e-06,
+      "loss": 0.138,
+      "recon_loss": 0.953905463218689,
+      "recon_trunc": 0.0,
+      "step": 1030,
+      "student_loss": 0.026124684140086174,
+      "teacher_loss": 0.006706951651722193
+    },
+    {
+      "epoch": 0.2983783783783784,
+      "grad_norm": 2.625,
+      "kd_loss": 0.014704967848956585,
+      "learning_rate": 2.32451311927106e-06,
+      "loss": 0.1245,
+      "recon_loss": 0.91652911901474,
+      "recon_trunc": 0.0,
+      "step": 1035,
+      "student_loss": 0.05515585094690323,
+      "teacher_loss": 0.049545880407094955
+    },
+    {
+      "epoch": 0.2998198198198198,
+      "grad_norm": 2.53125,
+      "kd_loss": 0.016647635027766228,
+      "learning_rate": 2.279068762183267e-06,
+      "loss": 0.1364,
+      "recon_loss": 0.7233182787895203,
+      "recon_trunc": 0.0,
+      "step": 1040,
+      "student_loss": 0.02766493707895279,
+      "teacher_loss": 0.028735877946019173
+    },
+    {
+      "epoch": 0.30126126126126124,
+      "grad_norm": 3.15625,
+      "kd_loss": 0.010429332964122295,
+      "learning_rate": 2.233941527638848e-06,
+      "loss": 0.1404,
+      "recon_loss": 0.10742680728435516,
+      "recon_trunc": 0.0,
+      "step": 1045,
+      "student_loss": 0.005210862960666418,
+      "teacher_loss": 0.0038872945588082075
+    },
+    {
+      "epoch": 0.3027027027027027,
+      "grad_norm": 2.71875,
+      "kd_loss": 0.02764248102903366,
+      "learning_rate": 2.1891366751839443e-06,
+      "loss": 0.1329,
+      "recon_loss": 0.7775788903236389,
+      "recon_trunc": 0.0,
+      "step": 1050,
+      "student_loss": 0.014028939418494701,
+      "teacher_loss": 0.022560402750968933
+    },
+    {
+      "epoch": 0.30414414414414415,
+      "grad_norm": 2.703125,
+      "kd_loss": 0.01924763061106205,
+      "learning_rate": 2.1446594267913003e-06,
+      "loss": 0.1199,
+      "recon_loss": 0.4747641682624817,
+      "recon_trunc": 0.0,
+      "step": 1055,
+      "student_loss": 0.008005080744624138,
+      "teacher_loss": 0.004270176403224468
+    },
+    {
+      "epoch": 0.3055855855855856,
+      "grad_norm": 3.4375,
+      "kd_loss": 0.01804213784635067,
+      "learning_rate": 2.1005149662516517e-06,
+      "loss": 0.1342,
+      "recon_loss": 0.3403450846672058,
+      "recon_trunc": 0.0,
+      "step": 1060,
+      "student_loss": 0.010722644627094269,
+      "teacher_loss": 0.006150628440082073
+    },
+    {
+      "epoch": 0.307027027027027,
+      "grad_norm": 1.9453125,
+      "kd_loss": 0.025220416486263275,
+      "learning_rate": 2.0567084385695475e-06,
+      "loss": 0.1427,
+      "recon_loss": 0.6151396036148071,
+      "recon_trunc": 0.0,
+      "step": 1065,
+      "student_loss": 0.018266428261995316,
+      "teacher_loss": 0.004366562236100435
+    },
+    {
+      "epoch": 0.3084684684684685,
+      "grad_norm": 2.625,
+      "kd_loss": 0.024269362911581993,
+      "learning_rate": 2.0132449493637095e-06,
+      "loss": 0.1262,
+      "recon_loss": 0.6272397637367249,
+      "recon_trunc": 0.029411764815449715,
+      "step": 1070,
+      "student_loss": 0.023181989789009094,
+      "teacher_loss": 0.018925365060567856
+    },
+    {
+      "epoch": 0.3099099099099099,
+      "grad_norm": 3.46875,
+      "kd_loss": 0.02885393612086773,
+      "learning_rate": 1.9701295642719836e-06,
+      "loss": 0.1356,
+      "recon_loss": 0.8617491126060486,
+      "recon_trunc": 0.0,
+      "step": 1075,
+      "student_loss": 0.034016359597444534,
+      "teacher_loss": 0.022477561607956886
+    },
+    {
+      "epoch": 0.31135135135135134,
+      "grad_norm": 4.40625,
+      "kd_loss": 0.02103048749268055,
+      "learning_rate": 1.927367308360926e-06,
+      "loss": 0.1467,
+      "recon_loss": 1.1568284034729004,
+      "recon_trunc": 0.0,
+      "step": 1080,
+      "student_loss": 0.003962129354476929,
+      "teacher_loss": 0.0014970928896218538
+    },
+    {
+      "epoch": 0.3127927927927928,
+      "grad_norm": 2.90625,
+      "kd_loss": 0.013153644278645515,
+      "learning_rate": 1.8849631655401557e-06,
+      "loss": 0.1376,
+      "recon_loss": 1.1626920700073242,
+      "recon_trunc": 0.0,
+      "step": 1085,
+      "student_loss": 0.02934510074555874,
+      "teacher_loss": 0.031375616788864136
+    },
+    {
+      "epoch": 0.31423423423423424,
+      "grad_norm": 2.09375,
+      "kd_loss": 0.018725693225860596,
+      "learning_rate": 1.8429220779814654e-06,
+      "loss": 0.1264,
+      "recon_loss": 0.5413309931755066,
+      "recon_trunc": 0.0,
+      "step": 1090,
+      "student_loss": 0.056059449911117554,
+      "teacher_loss": 0.03701705113053322
+    },
+    {
+      "epoch": 0.31567567567567567,
+      "grad_norm": 2.828125,
+      "kd_loss": 0.02253834344446659,
+      "learning_rate": 1.8012489455428261e-06,
+      "loss": 0.142,
+      "recon_loss": 0.6108108162879944,
+      "recon_trunc": 0.0,
+      "step": 1095,
+      "student_loss": 0.014023230411112309,
+      "teacher_loss": 0.0018756055505946279
+    },
+    {
+      "epoch": 0.3171171171171171,
+      "grad_norm": 2.484375,
+      "kd_loss": 0.03115518018603325,
+      "learning_rate": 1.7599486251973057e-06,
+      "loss": 0.1384,
+      "recon_loss": 1.480132818222046,
+      "recon_trunc": 0.125,
+      "step": 1100,
+      "student_loss": 0.021573340520262718,
+      "teacher_loss": 0.02037632279098034
+    },
+    {
+      "epoch": 0.3185585585585586,
+      "grad_norm": 1.96875,
+      "kd_loss": 0.02009745128452778,
+      "learning_rate": 1.7190259304670038e-06,
+      "loss": 0.12,
+      "recon_loss": 0.42386987805366516,
+      "recon_trunc": 0.0,
+      "step": 1105,
+      "student_loss": 0.02092079259455204,
+      "teacher_loss": 0.02166123501956463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.109375,
+      "kd_loss": 0.022645533084869385,
+      "learning_rate": 1.6784856308620217e-06,
+      "loss": 0.1385,
+      "recon_loss": 0.22908571362495422,
+      "recon_trunc": 0.0,
+      "step": 1110,
+      "student_loss": 0.015394425019621849,
+      "teacher_loss": 0.00312183634378016
+    },
+    {
+      "epoch": 0.32144144144144143,
+      "grad_norm": 2.328125,
+      "kd_loss": 0.014049198478460312,
+      "learning_rate": 1.638332451324593e-06,
+      "loss": 0.142,
+      "recon_loss": 1.068387508392334,
+      "recon_trunc": 0.0,
+      "step": 1115,
+      "student_loss": 0.009211467579007149,
+      "teacher_loss": 0.008737200871109962
+    },
+    {
+      "epoch": 0.32288288288288286,
+      "grad_norm": 2.453125,
+      "kd_loss": 0.007674572058022022,
+      "learning_rate": 1.5985710716783936e-06,
+      "loss": 0.1421,
+      "recon_loss": 1.193234920501709,
+      "recon_trunc": 0.0,
+      "step": 1120,
+      "student_loss": 0.003625984536483884,
+      "teacher_loss": 0.003345909994095564
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 2.59375,
+      "kd_loss": 0.007635142188519239,
+      "learning_rate": 1.5592061260831026e-06,
+      "loss": 0.1782,
+      "recon_loss": 4.4712934494018555,
+      "recon_trunc": 0.0,
+      "step": 1125,
+      "student_loss": 0.0413956344127655,
+      "teacher_loss": 0.04025491327047348
+    },
+    {
+      "epoch": 0.32576576576576577,
+      "grad_norm": 2.484375,
+      "kd_loss": 0.007928218692541122,
+      "learning_rate": 1.5202422024943069e-06,
+      "loss": 0.1217,
+      "recon_loss": 0.18418467044830322,
+      "recon_trunc": 0.0,
+      "step": 1130,
+      "student_loss": 0.0037010947708040476,
+      "teacher_loss": 0.0033041913993656635
+    },
+    {
+      "epoch": 0.3272072072072072,
+      "grad_norm": 2.703125,
+      "kd_loss": 0.026051659137010574,
+      "learning_rate": 1.4816838421287693e-06,
+      "loss": 0.1523,
+      "recon_loss": 0.7447088360786438,
+      "recon_trunc": 0.0,
+      "step": 1135,
+      "student_loss": 0.05436738207936287,
+      "teacher_loss": 0.025246668606996536
+    },
+    {
+      "epoch": 0.3286486486486486,
+      "grad_norm": 3.71875,
+      "kd_loss": 0.015343812294304371,
+      "learning_rate": 1.4435355389351553e-06,
+      "loss": 0.1348,
+      "recon_loss": 0.2703576385974884,
+      "recon_trunc": 0.0,
+      "step": 1140,
+      "student_loss": 0.007352327927947044,
+      "teacher_loss": 0.0035013502929359674
+    },
+    {
+      "epoch": 0.3300900900900901,
+      "grad_norm": 2.59375,
+      "kd_loss": 0.018555115908384323,
+      "learning_rate": 1.4058017390702671e-06,
+      "loss": 0.1584,
+      "recon_loss": 2.1659674644470215,
+      "recon_trunc": 0.0,
+      "step": 1145,
+      "student_loss": 0.03152637928724289,
+      "teacher_loss": 0.00556741189211607
+    },
+    {
+      "epoch": 0.33153153153153153,
+      "grad_norm": 2.265625,
+      "kd_loss": 0.012929409742355347,
+      "learning_rate": 1.368486840380851e-06,
+      "loss": 0.1394,
+      "recon_loss": 0.5746572613716125,
+      "recon_trunc": 0.0,
+      "step": 1150,
+      "student_loss": 0.019136246293783188,
+      "teacher_loss": 0.009017867036163807
+    },
+    {
+      "epoch": 0.33297297297297296,
+      "grad_norm": 2.65625,
+      "kd_loss": 0.013294035568833351,
+      "learning_rate": 1.3315951918910241e-06,
+      "loss": 0.1387,
+      "recon_loss": 1.553269863128662,
+      "recon_trunc": 0.0,
+      "step": 1155,
+      "student_loss": 0.0027628224343061447,
+      "teacher_loss": 0.00200835894793272
+    },
+    {
+      "epoch": 0.33441441441441444,
+      "grad_norm": 3.640625,
+      "kd_loss": 0.03295166790485382,
+      "learning_rate": 1.2951310932954024e-06,
+      "loss": 0.1298,
+      "recon_loss": 0.18797072768211365,
+      "recon_trunc": 0.0,
+      "step": 1160,
+      "student_loss": 0.031247230246663094,
+      "teacher_loss": 0.008808013051748276
+    },
+    {
+      "epoch": 0.33585585585585587,
+      "grad_norm": 3.890625,
+      "kd_loss": 0.025950822979211807,
+      "learning_rate": 1.2590987944579808e-06,
+      "loss": 0.1606,
+      "recon_loss": 0.5768255591392517,
+      "recon_trunc": 0.0,
+      "step": 1165,
+      "student_loss": 0.029426326975226402,
+      "teacher_loss": 0.013679691590368748
+    },
+    {
+      "epoch": 0.3372972972972973,
+      "grad_norm": 3.09375,
+      "kd_loss": 0.018202271312475204,
+      "learning_rate": 1.2235024949167994e-06,
+      "loss": 0.1448,
+      "recon_loss": 0.5929192900657654,
+      "recon_trunc": 0.0,
+      "step": 1170,
+      "student_loss": 0.025459816679358482,
+      "teacher_loss": 0.011100600473582745
+    },
+    {
+      "epoch": 0.3387387387387387,
+      "grad_norm": 2.15625,
+      "kd_loss": 0.030640222132205963,
+      "learning_rate": 1.1883463433945092e-06,
+      "loss": 0.1502,
+      "recon_loss": 0.2425113171339035,
+      "recon_trunc": 0.0,
+      "step": 1175,
+      "student_loss": 0.02755647525191307,
+      "teacher_loss": 0.010890517383813858
+    },
+    {
+      "epoch": 0.3401801801801802,
+      "grad_norm": 2.125,
+      "kd_loss": 0.025691065937280655,
+      "learning_rate": 1.1536344373148245e-06,
+      "loss": 0.1446,
+      "recon_loss": 0.7186999320983887,
+      "recon_trunc": 0.0,
+      "step": 1180,
+      "student_loss": 0.05646064504981041,
+      "teacher_loss": 0.01310134306550026
+    },
+    {
+      "epoch": 0.34162162162162163,
+      "grad_norm": 2.53125,
+      "kd_loss": 0.021015869453549385,
+      "learning_rate": 1.119370822324981e-06,
+      "loss": 0.1314,
+      "recon_loss": 0.40416237711906433,
+      "recon_trunc": 0.0,
+      "step": 1185,
+      "student_loss": 0.01777879148721695,
+      "teacher_loss": 0.012832175008952618
+    },
+    {
+      "epoch": 0.34306306306306306,
+      "grad_norm": 2.53125,
+      "kd_loss": 0.017006749287247658,
+      "learning_rate": 1.0855594918242157e-06,
+      "loss": 0.1395,
+      "recon_loss": 0.5203812718391418,
+      "recon_trunc": 0.0,
+      "step": 1190,
+      "student_loss": 0.03594272956252098,
+      "teacher_loss": 0.003274575574323535
+    },
+    {
+      "epoch": 0.3445045045045045,
+      "grad_norm": 2.875,
+      "kd_loss": 0.01839633099734783,
+      "learning_rate": 1.0522043864983428e-06,
+      "loss": 0.1273,
+      "recon_loss": 0.10688424855470657,
+      "recon_trunc": 0.0,
+      "step": 1195,
+      "student_loss": 0.004460033029317856,
+      "teacher_loss": 0.0046429503709077835
+    },
+    {
+      "epoch": 0.34594594594594597,
+      "grad_norm": 2.890625,
+      "kd_loss": 0.01323331892490387,
+      "learning_rate": 1.0193093938604627e-06,
+      "loss": 0.1464,
+      "recon_loss": 2.382429361343384,
+      "recon_trunc": 0.0,
+      "step": 1200,
+      "student_loss": 0.01494783814996481,
+      "teacher_loss": 0.004788065794855356
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deba083ea8ebf0ad6778aea83be6c94da22d25020a756643cc25829419a5ddc1
+size 5841

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/added_tokens.json

@@ -0,0 +1,35 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|action_sep|>": 151670,
+  "<|arg_sep|>": 151671,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|call_sep|>": 151666,
+  "<|end_of_text|>": 151673,
+  "<|endoftext|>": 151643,
+  "<|exception_sep|>": 151669,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|frame_sep|>": 151672,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|latent_end|>": 151675,
+  "<|latent_start|>": 151674,
+  "<|line_sep|>": 151667,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|return_sep|>": 151668,
+  "<|trace_context_start|>": 151665,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/config.json

@@ -0,0 +1,58 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b71bad11c3764860c40ef3f83d103299df6a7b43eabff70c20a9f22e89eb716e
+size 4539424633

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ab43c434fa541447c61abe3c477ca5d9b4b2bb921f17c93ec4e0226b6219611
+size 3097074235

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4aa03284661460f7588cf2ed51cae6f25c4941e23d7976dea3db0885dadb428
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97017e17d6838d8a09f84b48e28a5938a238a3e50e6a6671b6a2439f52a40fa6
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5a6d9e5e8614e5b7bcfddbab25b719e621f1ff67f400bc9a73dc5af42659928
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:841e9f5079dd6c6b4131e29e2e887ebf784990e17a6d8d8c7131ffb11723d069
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:475539e52c6e83f2f5a22f5cabb5de882acee72c74676095cdf00d332b3d8c01
+size 1465

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/thought_projector.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55f87785d58cc04afd926a54db9c0b16e667f818d6e16d273ed26afda9ccc027
+size 9445953

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83a790d654474f5dfe225f889afd0210313eb1083f942671f2c4b8e95a1c922b
+size 11424004

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/tokenizer_config.json

@@ -0,0 +1,295 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<|trace_context_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151666": {
+      "content": "<|call_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151667": {
+      "content": "<|line_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151668": {
+      "content": "<|return_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151669": {
+      "content": "<|exception_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151670": {
+      "content": "<|action_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151671": {
+      "content": "<|arg_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151672": {
+      "content": "<|frame_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151673": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151674": {
+      "content": "<|latent_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151675": {
+      "content": "<|latent_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-1500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deba083ea8ebf0ad6778aea83be6c94da22d25020a756643cc25829419a5ddc1
+size 5841

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/added_tokens.json

@@ -0,0 +1,35 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|action_sep|>": 151670,
+  "<|arg_sep|>": 151671,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|call_sep|>": 151666,
+  "<|end_of_text|>": 151673,
+  "<|endoftext|>": 151643,
+  "<|exception_sep|>": 151669,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|frame_sep|>": 151672,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|latent_end|>": 151675,
+  "<|latent_start|>": 151674,
+  "<|line_sep|>": 151667,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|return_sep|>": 151668,
+  "<|trace_context_start|>": 151665,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/config.json

@@ -0,0 +1,58 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfdd109a34a87ffe05a3d4aac1eb41490ef9d3f5803439dd642f1ee790f5307e
+size 4539424633

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dbacf71a7a7e12ba97d5ec2c85561003dbfe4542570bc5c05759ee00c480eed
+size 3097074235

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4aa03284661460f7588cf2ed51cae6f25c4941e23d7976dea3db0885dadb428
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97017e17d6838d8a09f84b48e28a5938a238a3e50e6a6671b6a2439f52a40fa6
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5a6d9e5e8614e5b7bcfddbab25b719e621f1ff67f400bc9a73dc5af42659928
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:841e9f5079dd6c6b4131e29e2e887ebf784990e17a6d8d8c7131ffb11723d069
+size 15429

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a1cf7dc915fd8683200c4cab003d16bce9119c6dac8f6d21b4ba406ae5f3bfc
+size 1465

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/codi_recon_delta_1.5b_ls1_rw0.1/checkpoint-300/thought_projector.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d4f754fd6bb5bb1688885b3fc7a9fb6aa7dd904a1e2b4f0028fbaacbdda017b
+size 9445953