Python4 AFT v2 adapter runs 20260813T154138Z mixed_1ep

.gitattributes

@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 runs/20260813T154138Z/arms/ordered_4ep/adapter/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 runs/20260813T154138Z/arms/control/adapter/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 runs/20260813T154138Z/arms/ordered_1ep/adapter/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+runs/20260813T154138Z/arms/mixed_1ep/adapter/tokenizer.json filter=lfs diff=lfs merge=lfs -text

runs/20260813T154138Z/arms/mixed_1ep/adapter/adapter_config.json

@@ -0,0 +1,231 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "/workspace/python4-aft-v2-state/20260813T154138Z/mixed_1ep/parent/dose_1ep_70m/sft/end",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 128,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "34.self_attn.k_proj",
+    "42.self_attn.v_proj",
+    "52.self_attn.q_proj",
+    "32.self_attn.k_proj",
+    "54.self_attn.v_proj",
+    "44.self_attn.k_proj",
+    "language_model.layers.8.self_attn.q_proj",
+    "60.self_attn.q_proj",
+    "language_model.layers.17.self_attn.v_proj",
+    "language_model.layers.9.self_attn.v_proj",
+    "36.self_attn.v_proj",
+    "35.self_attn.q_proj",
+    "58.self_attn.k_proj",
+    "56.self_attn.v_proj",
+    "language_model.layers.3.self_attn.k_proj",
+    "59.self_attn.k_proj",
+    "language_model.layers.13.self_attn.v_proj",
+    "language_model.layers.23.self_attn.v_proj",
+    "language_model.layers.16.self_attn.q_proj",
+    "language_model.layers.10.self_attn.v_proj",
+    "language_model.layers.11.self_attn.k_proj",
+    "language_model.layers.2.self_attn.k_proj",
+    "43.self_attn.q_proj",
+    "61.self_attn.q_proj",
+    "language_model.layers.15.self_attn.v_proj",
+    "29.self_attn.q_proj",
+    "38.self_attn.v_proj",
+    "27.self_attn.q_proj",
+    "language_model.layers.24.self_attn.q_proj",
+    "46.self_attn.q_proj",
+    "60.self_attn.k_proj",
+    "language_model.layers.26.self_attn.k_proj",
+    "language_model.layers.5.self_attn.q_proj",
+    "55.self_attn.k_proj",
+    "58.self_attn.q_proj",
+    "46.self_attn.v_proj",
+    "language_model.layers.4.self_attn.k_proj",
+    "language_model.layers.24.self_attn.k_proj",
+    "50.self_attn.k_proj",
+    "32.self_attn.q_proj",
+    "33.self_attn.q_proj",
+    "language_model.layers.15.self_attn.q_proj",
+    "41.self_attn.q_proj",
+    "language_model.layers.12.self_attn.v_proj",
+    "35.self_attn.v_proj",
+    "52.self_attn.v_proj",
+    "language_model.layers.15.self_attn.k_proj",
+    "language_model.layers.3.self_attn.v_proj",
+    "39.self_attn.v_proj",
+    "28.self_attn.q_proj",
+    "47.self_attn.v_proj",
+    "55.self_attn.q_proj",
+    "41.self_attn.v_proj",
+    "30.self_attn.q_proj",
+    "language_model.layers.4.self_attn.q_proj",
+    "49.self_attn.v_proj",
+    "language_model.layers.8.self_attn.k_proj",
+    "language_model.layers.11.self_attn.v_proj",
+    "48.self_attn.v_proj",
+    "language_model.layers.13.self_attn.q_proj",
+    "54.self_attn.q_proj",
+    "57.self_attn.v_proj",
+    "language_model.layers.1.self_attn.v_proj",
+    "language_model.layers.19.self_attn.q_proj",
+    "language_model.layers.6.self_attn.v_proj",
+    "language_model.layers.25.self_attn.q_proj",
+    "language_model.layers.0.self_attn.v_proj",
+    "38.self_attn.q_proj",
+    "language_model.layers.22.self_attn.v_proj",
+    "language_model.layers.10.self_attn.k_proj",
+    "language_model.layers.20.self_attn.k_proj",
+    "language_model.layers.20.self_attn.v_proj",
+    "up_proj",
+    "60.self_attn.v_proj",
+    "39.self_attn.k_proj",
+    "55.self_attn.v_proj",
+    "57.self_attn.k_proj",
+    "43.self_attn.v_proj",
+    "36.self_attn.q_proj",
+    "language_model.layers.13.self_attn.k_proj",
+    "language_model.layers.14.self_attn.q_proj",
+    "36.self_attn.k_proj",
+    "52.self_attn.k_proj",
+    "47.self_attn.k_proj",
+    "30.self_attn.k_proj",
+    "35.self_attn.k_proj",
+    "language_model.layers.20.self_attn.q_proj",
+    "language_model.layers.3.self_attn.q_proj",
+    "51.self_attn.q_proj",
+    "42.self_attn.q_proj",
+    "46.self_attn.k_proj",
+    "37.self_attn.k_proj",
+    "language_model.layers.11.self_attn.q_proj",
+    "27.self_attn.v_proj",
+    "language_model.layers.25.self_attn.v_proj",
+    "61.self_attn.v_proj",
+    "45.self_attn.k_proj",
+    "42.self_attn.k_proj",
+    "58.self_attn.v_proj",
+    "48.self_attn.k_proj",
+    "language_model.layers.5.self_attn.v_proj",
+    "50.self_attn.v_proj",
+    "language_model.layers.17.self_attn.q_proj",
+    "language_model.layers.23.self_attn.k_proj",
+    "29.self_attn.k_proj",
+    "31.self_attn.q_proj",
+    "language_model.layers.7.self_attn.q_proj",
+    "down_proj",
+    "language_model.layers.2.self_attn.v_proj",
+    "45.self_attn.q_proj",
+    "language_model.layers.14.self_attn.v_proj",
+    "o_proj",
+    "49.self_attn.q_proj",
+    "59.self_attn.q_proj",
+    "54.self_attn.k_proj",
+    "37.self_attn.q_proj",
+    "28.self_attn.k_proj",
+    "language_model.layers.19.self_attn.k_proj",
+    "33.self_attn.k_proj",
+    "28.self_attn.v_proj",
+    "language_model.layers.9.self_attn.q_proj",
+    "language_model.layers.5.self_attn.k_proj",
+    "language_model.layers.2.self_attn.q_proj",
+    "language_model.layers.26.self_attn.q_proj",
+    "51.self_attn.v_proj",
+    "41.self_attn.k_proj",
+    "61.self_attn.k_proj",
+    "language_model.layers.18.self_attn.v_proj",
+    "53.self_attn.k_proj",
+    "40.self_attn.k_proj",
+    "44.self_attn.q_proj",
+    "51.self_attn.k_proj",
+    "31.self_attn.v_proj",
+    "44.self_attn.v_proj",
+    "45.self_attn.v_proj",
+    "language_model.layers.6.self_attn.q_proj",
+    "27.self_attn.k_proj",
+    "56.self_attn.q_proj",
+    "language_model.layers.21.self_attn.q_proj",
+    "language_model.layers.21.self_attn.v_proj",
+    "language_model.layers.10.self_attn.q_proj",
+    "language_model.layers.9.self_attn.k_proj",
+    "language_model.layers.1.self_attn.k_proj",
+    "32.self_attn.v_proj",
+    "38.self_attn.k_proj",
+    "57.self_attn.q_proj",
+    "40.self_attn.v_proj",
+    "34.self_attn.q_proj",
+    "language_model.layers.4.self_attn.v_proj",
+    "language_model.layers.8.self_attn.v_proj",
+    "language_model.layers.26.self_attn.v_proj",
+    "language_model.layers.12.self_attn.q_proj",
+    "language_model.layers.7.self_attn.v_proj",
+    "gate_proj",
+    "language_model.layers.21.self_attn.k_proj",
+    "43.self_attn.k_proj",
+    "40.self_attn.q_proj",
+    "language_model.layers.16.self_attn.v_proj",
+    "language_model.layers.22.self_attn.q_proj",
+    "language_model.layers.19.self_attn.v_proj",
+    "34.self_attn.v_proj",
+    "50.self_attn.q_proj",
+    "56.self_attn.k_proj",
+    "language_model.layers.24.self_attn.v_proj",
+    "31.self_attn.k_proj",
+    "48.self_attn.q_proj",
+    "language_model.layers.0.self_attn.k_proj",
+    "59.self_attn.v_proj",
+    "33.self_attn.v_proj",
+    "language_model.layers.14.self_attn.k_proj",
+    "language_model.layers.25.self_attn.k_proj",
+    "language_model.layers.6.self_attn.k_proj",
+    "30.self_attn.v_proj",
+    "language_model.layers.0.self_attn.q_proj",
+    "language_model.layers.1.self_attn.q_proj",
+    "language_model.layers.17.self_attn.k_proj",
+    "53.self_attn.q_proj",
+    "language_model.layers.16.self_attn.k_proj",
+    "language_model.layers.7.self_attn.k_proj",
+    "language_model.layers.22.self_attn.k_proj",
+    "49.self_attn.k_proj",
+    "language_model.layers.18.self_attn.q_proj",
+    "language_model.layers.23.self_attn.q_proj",
+    "29.self_attn.v_proj",
+    "53.self_attn.v_proj",
+    "47.self_attn.q_proj",
+    "37.self_attn.v_proj",
+    "language_model.layers.12.self_attn.k_proj",
+    "39.self_attn.q_proj",
+    "language_model.layers.18.self_attn.k_proj"
+  ],
+  "target_parameters": [],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

runs/20260813T154138Z/arms/mixed_1ep/adapter/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c78608693af6fb7cd343777b05c2e01ae8abafe2726592280e5091dcb5b269d2
+size 1816396048

runs/20260813T154138Z/arms/mixed_1ep/adapter/chat_template.jinja

@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}

runs/20260813T154138Z/arms/mixed_1ep/adapter/config.json

@@ -0,0 +1,139 @@
+{
+  "architectures": [
+    "Gemma3ForConditionalGeneration"
+  ],
+  "boi_token_index": 255999,
+  "bos_token_id": 2,
+  "dtype": "bfloat16",
+  "eoi_token_index": 256000,
+  "eos_token_id": 106,
+  "image_token_index": 262144,
+  "initializer_range": 0.02,
+  "mm_tokens_per_image": 256,
+  "model_type": "gemma3",
+  "pad_token_id": 0,
+  "text_config": {
+    "_sliding_window_pattern": 6,
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attn_logit_softcapping": null,
+    "bos_token_id": 2,
+    "cache_implementation": "hybrid",
+    "dtype": "bfloat16",
+    "eos_token_id": 1,
+    "final_logit_softcapping": null,
+    "head_dim": 128,
+    "hidden_activation": "gelu_pytorch_tanh",
+    "hidden_size": 5376,
+    "initializer_range": 0.02,
+    "intermediate_size": 21504,
+    "layer_types": [
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention"
+    ],
+    "max_position_embeddings": 131072,
+    "model_type": "gemma3_text",
+    "num_attention_heads": 32,
+    "num_hidden_layers": 62,
+    "num_key_value_heads": 16,
+    "pad_token_id": 0,
+    "query_pre_attn_scalar": 168,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "full_attention": {
+        "factor": 8.0,
+        "rope_theta": 1000000.0,
+        "rope_type": "linear"
+      },
+      "sliding_attention": {
+        "rope_theta": 10000.0,
+        "rope_type": "default"
+      }
+    },
+    "sliding_window": 1024,
+    "sliding_window_pattern": 6,
+    "tie_word_embeddings": true,
+    "use_bidirectional_attention": false,
+    "use_cache": false,
+    "vocab_size": 262208
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "5.9.0",
+  "unsloth_fixed": true,
+  "use_cache": false,
+  "vision_config": {
+    "attention_dropout": 0.0,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "image_size": 896,
+    "intermediate_size": 4304,
+    "layer_norm_eps": 1e-06,
+    "model_type": "siglip_vision_model",
+    "num_attention_heads": 16,
+    "num_channels": 3,
+    "num_hidden_layers": 27,
+    "patch_size": 14,
+    "vision_use_head": false
+  }
+}

runs/20260813T154138Z/arms/mixed_1ep/adapter/processor_config.json

@@ -0,0 +1,28 @@
+{
+  "image_processor": {
+    "do_convert_rgb": null,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Gemma3ImageProcessor",
+    "image_seq_length": 256,
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "resample": 2,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "height": 896,
+      "width": 896
+    }
+  },
+  "image_seq_length": 256,
+  "processor_class": "Gemma3Processor"
+}

runs/20260813T154138Z/arms/mixed_1ep/adapter/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
+size 33384567

runs/20260813T154138Z/arms/mixed_1ep/adapter/tokenizer_config.json

@@ -0,0 +1,26 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<end_of_turn>",
+  "image_token": "<image_soft_token>",
+  "is_local": true,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 131072,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "padding_side": "left",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}