Add 11 files

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+training-metrics.png filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,99 @@
+---
+base_model: meta-llama/Llama-4-Scout-17B-16E-Instruct
+library_name: peft
+license: other
+tags:
+  - lora
+  - peft
+  - adapter
+  - adaption
+---
+
+# adaption_multi_agent_fraud_bench
+
+## Model Training
+
+A LORA adapter for `meta-llama/Llama-4-Scout-17B-16E-Instruct`. This model was trained with SFT using [Adaption](https://adaptionlabs.ai)'s AutoScientist on the multi_agent_fraud_bench dataset.
+
+
+![Training metrics](training-metrics.png)
+
+### AutoScientist Config
+
+```json
+{
+  "job_id": "76434061-b19d-478c-8444-b9665cf92c38",
+  "training_experiment_id": "c36f6473-f424-4138-ad1f-75db53d77e0c",
+  "original_model_name": "meta-llama/Llama-4-Scout-17B-16E-Instruct",
+  "trained_model_name": "adaption_multi_agent_fraud_bench",
+  "training_method": "sft",
+  "training_type": "lora",
+  "data_format": "chat",
+  "hyperparams": {
+    "lora": "true",
+    "lora_r": 64,
+    "n_evals": 5,
+    "n_epochs": 3,
+    "batch_size": "max",
+    "lora_alpha": 128,
+    "lora_dropout": 0,
+    "min_lr_ratio": 0.1,
+    "warmup_ratio": 0.1,
+    "weight_decay": 0.01,
+    "learning_rate": 0.0001,
+    "max_grad_norm": 1,
+    "base_model_size": "109B",
+    "train_on_inputs": "false",
+    "training_method": "sft",
+    "lr_scheduler_type": "cosine",
+    "scheduler_num_cycles": 0.5,
+    "lora_trainable_modules": "q_proj,k_proj,v_proj,o_proj,shared_expert.gate_proj,shared_expert.up_proj,shared_expert.down_proj,feed_forward.gate_proj,feed_forward.up_proj,feed_forward.down_proj"
+  }
+}
+```
+
+## Training Data
+
+The model was trained on 8,114 rows of adapted data with the following domain distribution: governance (35%), legal (30%), academic-education (15%), science (6%), marketing (5%), corporate-business (3%), medical (2%), personal-finance (1%), technology (1%), market-analysis (0%), writing-editing-communication (0%), architecture-design (0%), other (0%), career-workplace (0%), language (0%), hr (0%), logic (0%), culture (0%), religion (0%), roleplay (0%), news (0%), code (0%), dating (0%), how-to (0%), history (0%), data-analysis-visualization (0%).
+
+## Model Evaluation
+
+The model was evaluated on an in-distribution held-out test set as well as a broader domain-specific test set to measure generalization.
+
+
+![Win rates](win-rates.png)
+
+
+## How to use
+
+```bash
+pip install torch transformers peft
+```
+
+```python
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from peft import PeftModel
+
+BASE = "meta-llama/Llama-4-Scout-17B-16E-Instruct"
+ADAPTER = "<this-repo-id>"
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+dtype = torch.float32 if device == "cpu" else torch.bfloat16
+
+base = AutoModelForCausalLM.from_pretrained(BASE, dtype=dtype).to(device)
+model = PeftModel.from_pretrained(base, ADAPTER)
+# Optional: merge the LoRA weights into the base for faster inference
+model = model.merge_and_unload()
+model.eval()
+
+tokenizer = AutoTokenizer.from_pretrained(BASE)
+messages = [{"role": "user", "content": "Hello!"}]
+text = tokenizer.apply_chat_template(
+    messages, tokenize=False, add_generation_prompt=True)
+inputs = tokenizer(text, return_tensors="pt").to(device)
+
+with torch.inference_mode():
+    out = model.generate(**inputs, max_new_tokens=512)
+print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
+```

adapter_config.json

@@ -0,0 +1,42 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "togethercomputer/Llama-4-Scout-17B-16E-Instruct_bnb_4bit",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": [],
+  "fan_in_fan_out": false,
+  "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,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "o_proj",
+    "v_proj",
+    "feed_forward.down_proj",
+    "shared_expert.gate_proj",
+    "feed_forward.gate_proj",
+    "k_proj",
+    "feed_forward.up_proj",
+    "shared_expert.down_proj",
+    "shared_expert.up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d875eec5b10b03b63078dd13749ee104225af7add97b06c8a6741ddd16db8921
+size 893484120

chat_template.jinja

@@ -0,0 +1,123 @@
+{{- bos_token }}
+{%- if custom_tools is defined %}
+    {%- set tools = custom_tools %}
+{%- endif %}
+{%- if not tools_in_user_message is defined %}
+    {%- set tools_in_user_message = true %}
+{%- endif %}
+{%- if not date_string is defined %}
+    {%- if strftime_now is defined %}
+        {%- set date_string = strftime_now("%d %b %Y") %}
+    {%- else %}
+        {%- set date_string = "26 Jul 2024" %}
+    {%- endif %}
+{%- endif %}
+{%- if not tools is defined %}
+    {%- set tools = none %}
+{%- endif %}
+
+{#- This block extracts the system message, so we can slot it into the right place. #}
+{%- if messages[0]['role'] == 'system' %}    
+    {%- if messages[0]['content'] is string %}
+        {%- set system_message = messages[0]['content']|trim %}
+    {%- else %}
+        {#- FIXME: The processor requires an array, always. #}
+        {%- set system_message = messages[0]['content'][0]['text']|trim %}
+    {%- endif %}
+    {%- set messages = messages[1:] %}
+    {%- set user_supplied_system_message = true %}
+{%- else %}
+    {%- set system_message = "" %}
+    {%- set user_supplied_system_message = false %}
+{%- endif %}
+
+{#- System message if the user supplied one #}
+{%- if user_supplied_system_message %}
+    {{- "<|header_start|>system<|header_end|>\n\n" }}
+    {%- if tools is not none %}
+        {{- "Environment: ipython\n" }}
+    {%- endif %}
+    {%- if tools is not none and not tools_in_user_message %}
+        {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }}
+        {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+        {{- "Do not use variables.\n\n" }}
+        {%- for t in tools %}
+            {{- t | tojson(indent=4) }}
+            {{- "\n\n" }}
+        {%- endfor %}
+    {%- endif %}
+    {{- system_message }}
+    {{- "<|eot|>" }}
+{%- endif %}
+
+{#- Custom tools are passed in a user message with some extra guidance #}
+{%- if tools_in_user_message and not tools is none %}
+    {#- Extract the first user message so we can plug it in here #}
+    {%- if messages | length != 0 %}
+        {%- set first_user_message = messages[0]['content']|trim %}
+        {%- set messages = messages[1:] %}
+    {%- else %}
+        {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }}
+{%- endif %}
+    {{- '<|header_start|>user<|header_end|>\n\n' -}}
+    {{- "Given the following functions, please respond with a JSON for a function call " }}
+    {{- "with its proper arguments that best answers the given prompt.\n\n" }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+    {{- first_user_message + "<|eot|>"}}
+{%- endif %}
+
+{%- for message in messages %}
+    {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}
+    {{- '<|header_start|>' + message['role'] + '<|header_end|>\n\n' }}
+        {%- if message['content'] is string %}
+            {{- message['content'] }}
+        {%- else %}
+            {%- for content in message['content'] %}
+                {%- if content['type'] == 'image' %}
+                    {{- '<|image|>' }}
+                {%- elif content['type'] == 'text' %}
+                    {{- content['text'] }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- "<|eot|>" }}
+    {%- elif 'tool_calls' in message and message.tool_calls|length > 0 %}
+       {{- '<|header_start|>assistant<|header_end|>\n\n' -}}
+       {{- '<|python_start|>' }}
+        {%- if message['content'] is string %}
+            {{- message['content'] }}
+        {%- else %}
+            {%- for content in message['content'] %}
+                {%- if content['type'] == 'image' %}
+                    {{- '<|image|>' }}
+                {%- elif content['type'] == 'text' %}
+                    {{- content['text'] }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+       {{- '<|python_end|>' }}
+        {%- for tool_call in message.tool_calls %}
+           {{- '{"name": "' + tool_call.function.name + '", ' }}
+           {{- '"parameters": ' }}
+           {{- tool_call.function.arguments | tojson }}
+           {{- "}" }}
+        {%- endfor %}
+       {{- "<|eot|>" }}
+    {%- elif message.role == "tool" or message.role == "ipython" %}
+        {{- "<|header_start|>ipython<|header_end|>\n\n" }}
+        {%- if message.content is mapping or message.content is iterable %}
+            {{- message.content | tojson }}
+        {%- else %}
+            {{- message.content }}
+        {%- endif %}
+        {{- "<|eot|>" }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|header_start|>assistant<|header_end|>\n\n' }}
+{%- endif %}

config.json

@@ -0,0 +1,215 @@
+{
+  "architectures": [
+    "Llama4ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_chunk_size": 8192,
+  "attention_dropout": 0.0,
+  "attn_scale": 0.1,
+  "attn_temperature_tuning": true,
+  "bos_token_id": 200000,
+  "cache_implementation": "hybrid",
+  "dtype": "bfloat16",
+  "eos_token_id": 200008,
+  "floor_scale": 8192,
+  "for_llm_compressor": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "interleave_moe_layer_step": 1,
+  "intermediate_size": 8192,
+  "intermediate_size_mlp": 16384,
+  "layer_types": [
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "chunked_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 10485760,
+  "model_type": "llama4_text",
+  "moe_layers": [
+    0,
+    1,
+    2,
+    3,
+    4,
+    5,
+    6,
+    7,
+    8,
+    9,
+    10,
+    11,
+    12,
+    13,
+    14,
+    15,
+    16,
+    17,
+    18,
+    19,
+    20,
+    21,
+    22,
+    23,
+    24,
+    25,
+    26,
+    27,
+    28,
+    29,
+    30,
+    31,
+    32,
+    33,
+    34,
+    35,
+    36,
+    37,
+    38,
+    39,
+    40,
+    41,
+    42,
+    43,
+    44,
+    45,
+    46,
+    47
+  ],
+  "no_rope_layer_interval": 4,
+  "no_rope_layers": [
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0,
+    1,
+    1,
+    1,
+    0
+  ],
+  "num_attention_heads": 40,
+  "num_experts_per_tok": 1,
+  "num_hidden_layers": 48,
+  "num_key_value_heads": 8,
+  "num_local_experts": 16,
+  "output_router_logits": false,
+  "pad_token_id": 200018,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "bfloat16",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_parameters": {
+    "factor": 16.0,
+    "high_freq_factor": 1.0,
+    "low_freq_factor": 1.0,
+    "original_max_position_embeddings": 8192,
+    "rope_theta": 500000.0,
+    "rope_type": "llama3"
+  },
+  "router_aux_loss_coef": 0.001,
+  "router_jitter_noise": 0.0,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.13.0",
+  "use_cache": false,
+  "use_qk_norm": true,
+  "vocab_size": 202048,
+  "torch_dtype": "bfloat16"
+}

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "bos_token": "<|begin_of_text|>",
+  "eos_token": "<|eot|>",
+  "pad_token": "<|finetune_right_pad|>"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:172c9eb4beafc72601690da3ccfcede5c2e6806a8d5ec1fca33e22acea8023a4
+size 27948578

tokenizer_config.json

@@ -0,0 +1,17 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|eot|>",
+  "is_local": false,
+  "local_files_only": true,
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 10485760,
+  "pad_token": "<|finetune_right_pad|>",
+  "padding_side": "right",
+  "processor_class": "Llama4Processor",
+  "tokenizer_class": "TokenizersBackend"
+}

trainer_state.json

@@ -0,0 +1,515 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 12,
+  "global_step": 63,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 7.683346271514893,
+      "learning_rate": 0.0,
+      "loss": 2.7236328125,
+      "step": 1
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 8.30794906616211,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 2.80615234375,
+      "step": 2
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 8.94153881072998,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 3.1728515625,
+      "step": 3
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 7.938409328460693,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 2.9375,
+      "step": 4
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 1.4468631744384766,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 1.834716796875,
+      "step": 5
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.7402069568634033,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 1.4140625,
+      "step": 6
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.0219361782073975,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 1.1376953125,
+      "step": 7
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.2759617567062378,
+      "learning_rate": 0.0001,
+      "loss": 0.978515625,
+      "step": 8
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.3914451599121094,
+      "learning_rate": 9.992920667580176e-05,
+      "loss": 0.91015625,
+      "step": 9
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 1.0541131496429443,
+      "learning_rate": 9.971704944519594e-05,
+      "loss": 0.894287109375,
+      "step": 10
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.7295523285865784,
+      "learning_rate": 9.936419583332062e-05,
+      "loss": 0.747314453125,
+      "step": 11
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.9586805701255798,
+      "learning_rate": 9.887175604818206e-05,
+      "loss": 0.73583984375,
+      "step": 12
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.9399129748344421,
+      "learning_rate": 9.82412794875295e-05,
+      "loss": 0.91162109375,
+      "step": 13
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5930038690567017,
+      "learning_rate": 9.747474986387654e-05,
+      "loss": 0.665283203125,
+      "step": 14
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.6943517327308655,
+      "learning_rate": 9.657457896300791e-05,
+      "loss": 0.68115234375,
+      "step": 15
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "eval_loss": 1.125,
+      "eval_runtime": 3.7655,
+      "eval_samples_per_second": 0.266,
+      "eval_steps_per_second": 0.266,
+      "step": 15
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.6625951528549194,
+      "learning_rate": 9.554359905560886e-05,
+      "loss": 0.820556640625,
+      "step": 16
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.7600706219673157,
+      "learning_rate": 9.438505398589392e-05,
+      "loss": 0.614013671875,
+      "step": 17
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.5712336301803589,
+      "learning_rate": 9.310258896527278e-05,
+      "loss": 0.68603515625,
+      "step": 18
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.6201907992362976,
+      "learning_rate": 9.17002391031667e-05,
+      "loss": 0.6240234375,
+      "step": 19
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.6295406818389893,
+      "learning_rate": 9.018241671106134e-05,
+      "loss": 0.636962890625,
+      "step": 20
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7266039252281189,
+      "learning_rate": 8.855389741974244e-05,
+      "loss": 0.641845703125,
+      "step": 21
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 0.6056920289993286,
+      "learning_rate": 8.681980515339464e-05,
+      "loss": 0.50732421875,
+      "step": 22
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.5053836107254028,
+      "learning_rate": 8.498559600784018e-05,
+      "loss": 0.4849853515625,
+      "step": 23
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.5425382256507874,
+      "learning_rate": 8.305704108364301e-05,
+      "loss": 0.560791015625,
+      "step": 24
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.6988992691040039,
+      "learning_rate": 8.104020832809127e-05,
+      "loss": 0.66650390625,
+      "step": 25
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 0.5177556872367859,
+      "learning_rate": 7.894144344319014e-05,
+      "loss": 1.3883743286132812,
+      "step": 26
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.6992603540420532,
+      "learning_rate": 7.67673499197358e-05,
+      "loss": 0.667724609375,
+      "step": 27
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "eval_loss": 1.0234375,
+      "eval_runtime": 3.7364,
+      "eval_samples_per_second": 0.268,
+      "eval_steps_per_second": 0.268,
+      "step": 27
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.5518877506256104,
+      "learning_rate": 7.452476826029011e-05,
+      "loss": 0.6201171875,
+      "step": 28
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.48834937810897827,
+      "learning_rate": 7.222075445642904e-05,
+      "loss": 0.59765625,
+      "step": 29
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.7777575254440308,
+      "learning_rate": 6.986255778798253e-05,
+      "loss": 0.549560546875,
+      "step": 30
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.5613299012184143,
+      "learning_rate": 6.745759801411822e-05,
+      "loss": 0.591552734375,
+      "step": 31
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.48722875118255615,
+      "learning_rate": 6.501344202803414e-05,
+      "loss": 0.4981689453125,
+      "step": 32
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.45788365602493286,
+      "learning_rate": 6.253778004871315e-05,
+      "loss": 0.47705078125,
+      "step": 33
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 0.5666241645812988,
+      "learning_rate": 6.003840142464886e-05,
+      "loss": 0.628173828125,
+      "step": 34
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.49418237805366516,
+      "learning_rate": 5.752317012567363e-05,
+      "loss": 0.474609375,
+      "step": 35
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.6069824695587158,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 0.467041015625,
+      "step": 36
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.5141736268997192,
+      "learning_rate": 5.247682987432637e-05,
+      "loss": 0.596435546875,
+      "step": 37
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.49380624294281006,
+      "learning_rate": 4.9961598575351155e-05,
+      "loss": 0.4202880859375,
+      "step": 38
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.48107779026031494,
+      "learning_rate": 4.7462219951286867e-05,
+      "loss": 0.49169921875,
+      "step": 39
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "eval_loss": 0.99609375,
+      "eval_runtime": 3.7334,
+      "eval_samples_per_second": 0.268,
+      "eval_steps_per_second": 0.268,
+      "step": 39
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.4944271743297577,
+      "learning_rate": 4.498655797196586e-05,
+      "loss": 0.435791015625,
+      "step": 40
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.46663668751716614,
+      "learning_rate": 4.2542401985881784e-05,
+      "loss": 0.45556640625,
+      "step": 41
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5421565175056458,
+      "learning_rate": 4.01374422120175e-05,
+      "loss": 0.4136962890625,
+      "step": 42
+    },
+    {
+      "epoch": 2.0476190476190474,
+      "grad_norm": 0.47634458541870117,
+      "learning_rate": 3.777924554357096e-05,
+      "loss": 0.3116455078125,
+      "step": 43
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.4998021125793457,
+      "learning_rate": 3.547523173970989e-05,
+      "loss": 0.36248779296875,
+      "step": 44
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.6161043643951416,
+      "learning_rate": 3.323265008026421e-05,
+      "loss": 0.395263671875,
+      "step": 45
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.5737324357032776,
+      "learning_rate": 3.105855655680986e-05,
+      "loss": 0.4962158203125,
+      "step": 46
+    },
+    {
+      "epoch": 2.238095238095238,
+      "grad_norm": 0.45135700702667236,
+      "learning_rate": 2.8959791671908742e-05,
+      "loss": 1.2419204711914062,
+      "step": 47
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.6822405457496643,
+      "learning_rate": 2.6942958916356998e-05,
+      "loss": 0.49658203125,
+      "step": 48
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.6327748894691467,
+      "learning_rate": 2.501440399215983e-05,
+      "loss": 0.4644775390625,
+      "step": 49
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.5168265104293823,
+      "learning_rate": 2.3180194846605367e-05,
+      "loss": 0.458740234375,
+      "step": 50
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.5705485939979553,
+      "learning_rate": 2.144610258025755e-05,
+      "loss": 0.402587890625,
+      "step": 51
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "eval_loss": 0.99609375,
+      "eval_runtime": 3.7308,
+      "eval_samples_per_second": 0.268,
+      "eval_steps_per_second": 0.268,
+      "step": 51
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 0.616422176361084,
+      "learning_rate": 1.981758328893866e-05,
+      "loss": 0.4619140625,
+      "step": 52
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 0.5285511612892151,
+      "learning_rate": 1.8299760896833295e-05,
+      "loss": 0.39013671875,
+      "step": 53
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.5321851968765259,
+      "learning_rate": 1.6897411034727218e-05,
+      "loss": 0.36767578125,
+      "step": 54
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.5941992402076721,
+      "learning_rate": 1.5614946014106085e-05,
+      "loss": 0.4990234375,
+      "step": 55
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.5202521085739136,
+      "learning_rate": 1.4456400944391146e-05,
+      "loss": 0.3797607421875,
+      "step": 56
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.6072654724121094,
+      "learning_rate": 1.3425421036992098e-05,
+      "loss": 0.375,
+      "step": 57
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 0.6598243713378906,
+      "learning_rate": 1.252525013612346e-05,
+      "loss": 0.496337890625,
+      "step": 58
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 0.5702089071273804,
+      "learning_rate": 1.1758720512470523e-05,
+      "loss": 0.3341064453125,
+      "step": 59
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.550778329372406,
+      "learning_rate": 1.1128243951817937e-05,
+      "loss": 0.404296875,
+      "step": 60
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 0.5884420871734619,
+      "learning_rate": 1.0635804166679386e-05,
+      "loss": 0.3558349609375,
+      "step": 61
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.5809072256088257,
+      "learning_rate": 1.0282950554804085e-05,
+      "loss": 0.376708984375,
+      "step": 62
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6065093874931335,
+      "learning_rate": 1.0070793324198253e-05,
+      "loss": 0.31134033203125,
+      "step": 63
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.984375,
+      "eval_runtime": 3.7577,
+      "eval_samples_per_second": 0.266,
+      "eval_steps_per_second": 0.266,
+      "step": 63
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 63,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0104242682923581e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}