Upload full trainer checkpoint step=141 (profile=llama3_2_1b_base_grpo, run_type=grpo)

.gitattributes

@@ -87,3 +87,4 @@ meta-llama__llama-3.2-3b/sft/checkpoints/checkpoint-29325/tokenizer.json filter=
 meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-1/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-12/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-24/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/tokenizer.json filter=lfs diff=lfs merge=lfs -text

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/README.md

@@ -0,0 +1,210 @@
+---
+base_model: unsloth/Llama-3.2-1B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/Llama-3.2-1B
+- grpo
+- lora
+- transformers
+- trl
+- unsloth
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.1

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/adapter_config.json

@@ -0,0 +1,50 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "LlamaForCausalLM",
+    "parent_library": "transformers.models.llama.modeling_llama",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/Llama-3.2-1B",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "gate_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0e2dfa0bd6aeae8ec5364292c5d8ab1899d4a4d88c63c04f9a7f580320bf1d4
+size 45118424

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/chat_template.jinja

@@ -0,0 +1,12 @@
+{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ messages[0]['content'] + '
+
+' }}{% set loop_messages = messages[1:] %}{% else %}{{ '' + '
+
+' }}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['role'] == 'user' %}{{ '### Instruction:
+' + message['content'] + '
+
+' }}{% elif message['role'] == 'assistant' %}{{ '### Response:
+' + message['content'] + eos_token + '
+
+' }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '### Response:
+' }}{% endif %}

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55d22fd9c17f1464bba6b1f170e34170f137f1bd6c530329e9e78d88c6f3ad09
+size 23161995

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/resume_sft_best/adapter_config.json

@@ -0,0 +1,50 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "LlamaForCausalLM",
+    "parent_library": "transformers.models.llama.modeling_llama",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/Llama-3.2-1B",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "v_proj",
+    "k_proj",
+    "down_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/resume_sft_best/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f362eb408190d63eced05c41115e5583583d1c850f3926290fc41437923cb6f2
+size 45118424

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b367e2de5a7f22fe9c96846f01a41c4f25cc0de7011662e71802a15be75c0cb
+size 14645

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1009768ff6ddd77a5c52776b75f8a04c7ddcc4367a6c1125e054ac6e250c2c1
+size 1465

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a65c6c5f9764771aa485e6a1f5e63d7d9af8477fe0777148c17476ecb2e09a05
+size 17210099

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/tokenizer_config.json

@@ -0,0 +1,18 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|end_of_text|>",
+  "from_slow": true,
+  "is_local": false,
+  "legacy": false,
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 131072,
+  "pad_token": "<|finetune_right_pad_id|>",
+  "padding_side": "right",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": null
+}

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.0007513428254753176,
+  "eval_steps": 500,
+  "global_step": 141,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 70.975,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 124.5,
+      "completions/max_terminated_length": 124.5,
+      "completions/mean_length": 64.078125,
+      "completions/mean_terminated_length": 64.078125,
+      "completions/min_length": 29.5,
+      "completions/min_terminated_length": 29.5,
+      "epoch": 5.328672521101543e-05,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 0.0,
+      "kl": 0.0,
+      "learning_rate": 1.4986013054482483e-08,
+      "loss": 0.003790590539574623,
+      "num_tokens": 14625.0,
+      "reward": -7.450580596923828e-09,
+      "reward_std": 0.3402964845299721,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -7.450580596923828e-09,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 10
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.00041666668839752675,
+      "clip_ratio/low_min": 0.00041666668839752675,
+      "clip_ratio/region_mean": 0.00041666668839752675,
+      "completion_length": 62.875,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 119.0,
+      "completions/max_terminated_length": 119.0,
+      "completions/mean_length": 79.65625,
+      "completions/mean_terminated_length": 79.65625,
+      "completions/min_length": 48.0,
+      "completions/min_terminated_length": 48.0,
+      "epoch": 0.00010657345042203086,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 0.0010868345852941275,
+      "kl": 0.0,
+      "learning_rate": 3.163713867057413e-08,
+      "loss": -0.007658571749925613,
+      "num_tokens": 23710.0,
+      "reward": -2.9802322387695312e-08,
+      "reward_std": 0.24734649062156677,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -2.9802322387695312e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160011053085327,
+      "step": 20
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 80.625,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 109.0,
+      "completions/max_terminated_length": 109.0,
+      "completions/mean_length": 79.46875,
+      "completions/mean_terminated_length": 79.46875,
+      "completions/min_length": 24.0,
+      "completions/min_terminated_length": 24.0,
+      "epoch": 0.0001598601756330463,
+      "frac_reward_zero_std": 0.75,
+      "grad_norm": 0.014402221888303757,
+      "kl": 0.0,
+      "learning_rate": 4.828826428666578e-08,
+      "loss": 0.004134796187281608,
+      "num_tokens": 32613.0,
+      "reward": 0.0,
+      "reward_std": 0.33609244227409363,
+      "rewards/TranscriptCorrectionGrpoReward/mean": 0.0,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 30
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 62.25,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 81.0,
+      "completions/max_terminated_length": 81.0,
+      "completions/mean_length": 58.625,
+      "completions/mean_terminated_length": 58.625,
+      "completions/min_length": 40.0,
+      "completions/min_terminated_length": 40.0,
+      "epoch": 0.00021314690084406172,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 0.005918669048696756,
+      "kl": 0.0,
+      "learning_rate": 6.493938990275742e-08,
+      "loss": 0.0006381096318364143,
+      "num_tokens": 39377.0,
+      "reward": 0.0,
+      "reward_std": 0.30376893281936646,
+      "rewards/TranscriptCorrectionGrpoReward/mean": 0.0,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 40
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 63.05,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 115.0,
+      "completions/max_terminated_length": 115.0,
+      "completions/mean_length": 65.03125,
+      "completions/mean_terminated_length": 65.03125,
+      "completions/min_length": 44.5,
+      "completions/min_terminated_length": 44.5,
+      "epoch": 0.00026643362605507714,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 0.0046983626671135426,
+      "kl": 0.0,
+      "learning_rate": 8.159051551884909e-08,
+      "loss": -0.006434406340122223,
+      "num_tokens": 54359.0,
+      "reward": 0.0,
+      "reward_std": 0.14104405045509338,
+      "rewards/TranscriptCorrectionGrpoReward/mean": 0.0,
+      "rewards/TranscriptCorrectionGrpoReward/std": 0.5080005526542664,
+      "step": 50
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0008283828385174275,
+      "clip_ratio/low_min": 0.0008283828385174275,
+      "clip_ratio/region_mean": 0.0008283828385174275,
+      "completion_length": 66.15,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 96.0,
+      "completions/max_terminated_length": 96.0,
+      "completions/mean_length": 61.625,
+      "completions/mean_terminated_length": 61.625,
+      "completions/min_length": 20.0,
+      "completions/min_terminated_length": 20.0,
+      "epoch": 0.0003197203512660926,
+      "frac_reward_zero_std": 0.875,
+      "grad_norm": 0.0,
+      "kl": 0.0,
+      "learning_rate": 9.824164113494073e-08,
+      "loss": 0.014849747717380523,
+      "num_tokens": 61423.0,
+      "reward": -1.4901161193847656e-08,
+      "reward_std": 0.29814237356185913,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -1.4901161193847656e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 60
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.00038759689778089523,
+      "clip_ratio/low_min": 0.00038759689778089523,
+      "clip_ratio/region_mean": 0.00038759689778089523,
+      "completion_length": 60.125,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 129.0,
+      "completions/max_terminated_length": 129.0,
+      "completions/mean_length": 60.625,
+      "completions/mean_terminated_length": 60.625,
+      "completions/min_length": 28.0,
+      "completions/min_terminated_length": 28.0,
+      "epoch": 0.00037300707647710803,
+      "frac_reward_zero_std": 0.875,
+      "grad_norm": 0.0,
+      "kl": 0.0,
+      "learning_rate": 1.1489276675103237e-07,
+      "loss": -0.008947646617889405,
+      "num_tokens": 68375.0,
+      "reward": -1.4901161193847656e-08,
+      "reward_std": 0.03345904499292374,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -1.4901161193847656e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160011053085327,
+      "step": 70
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 74.05,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 112.0,
+      "completions/max_terminated_length": 112.0,
+      "completions/mean_length": 76.90625,
+      "completions/mean_terminated_length": 76.90625,
+      "completions/min_length": 30.0,
+      "completions/min_terminated_length": 30.0,
+      "epoch": 0.00042629380168812344,
+      "frac_reward_zero_std": 0.375,
+      "grad_norm": 0.02389357052743435,
+      "kl": 0.0,
+      "learning_rate": 1.3154389236712402e-07,
+      "loss": -0.0001737307757139206,
+      "num_tokens": 76956.0,
+      "reward": -1.4901161193847656e-08,
+      "reward_std": 0.6354520320892334,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -1.4901161193847656e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 80
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 70.475,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 98.0,
+      "completions/max_terminated_length": 98.0,
+      "completions/mean_length": 70.0625,
+      "completions/mean_terminated_length": 70.0625,
+      "completions/min_length": 37.5,
+      "completions/min_terminated_length": 37.5,
+      "epoch": 0.00047958052689913886,
+      "frac_reward_zero_std": 0.875,
+      "grad_norm": 0.0,
+      "kl": 0.0,
+      "learning_rate": 1.4819501798321567e-07,
+      "loss": -0.00633343905210495,
+      "num_tokens": 92712.0,
+      "reward": -7.450580596923828e-09,
+      "reward_std": 0.254583939909935,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -7.450580596923828e-09,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 90
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 73.45,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 98.0,
+      "completions/max_terminated_length": 98.0,
+      "completions/mean_length": 79.03125,
+      "completions/mean_terminated_length": 79.03125,
+      "completions/min_length": 56.0,
+      "completions/min_terminated_length": 56.0,
+      "epoch": 0.0005328672521101543,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 0.04292331635951996,
+      "kl": 0.0,
+      "learning_rate": 1.6484614359930732e-07,
+      "loss": 0.0025398530066013335,
+      "num_tokens": 101453.0,
+      "reward": 2.2351741790771484e-08,
+      "reward_std": 0.4319751560688019,
+      "rewards/TranscriptCorrectionGrpoReward/mean": 2.2351741790771484e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 100
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 69.1,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 87.0,
+      "completions/max_terminated_length": 87.0,
+      "completions/mean_length": 64.21875,
+      "completions/mean_terminated_length": 64.21875,
+      "completions/min_length": 42.0,
+      "completions/min_terminated_length": 42.0,
+      "epoch": 0.0005861539773211698,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 0.015379291027784348,
+      "kl": 0.0,
+      "learning_rate": 1.8149726921539897e-07,
+      "loss": -0.016414041817188262,
+      "num_tokens": 108716.0,
+      "reward": -1.4901161193847656e-08,
+      "reward_std": 0.4683850407600403,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -1.4901161193847656e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 110
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 62.675,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 100.0,
+      "completions/max_terminated_length": 100.0,
+      "completions/mean_length": 61.25,
+      "completions/mean_terminated_length": 61.25,
+      "completions/min_length": 31.0,
+      "completions/min_terminated_length": 31.0,
+      "epoch": 0.0006394407025321852,
+      "frac_reward_zero_std": 0.875,
+      "grad_norm": 0.018463263288140297,
+      "kl": 0.0,
+      "learning_rate": 1.9814839483149064e-07,
+      "loss": 0.01983743757009506,
+      "num_tokens": 115916.0,
+      "reward": 0.0,
+      "reward_std": 0.10585122555494308,
+      "rewards/TranscriptCorrectionGrpoReward/mean": 0.0,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 120
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completion_length": 53.6,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 106.0,
+      "completions/max_terminated_length": 106.0,
+      "completions/mean_length": 62.171875,
+      "completions/mean_terminated_length": 62.171875,
+      "completions/min_length": 31.0,
+      "completions/min_terminated_length": 31.0,
+      "epoch": 0.0006927274277432006,
+      "frac_reward_zero_std": 0.8125,
+      "grad_norm": 0.0817682221531868,
+      "kl": 0.0,
+      "learning_rate": 2.147995204475823e-07,
+      "loss": 0.003151830658316612,
+      "num_tokens": 130307.0,
+      "reward": -7.450580596923828e-09,
+      "reward_std": 0.32130110263824463,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -7.450580596923828e-09,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 130
+    },
+    {
+      "clip_ratio/high_max": 0.0003546099178493023,
+      "clip_ratio/high_mean": 0.0003546099178493023,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0003546099178493023,
+      "completion_length": 78.425,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 124.0,
+      "completions/max_terminated_length": 124.0,
+      "completions/mean_length": 76.875,
+      "completions/mean_terminated_length": 76.875,
+      "completions/min_length": 36.0,
+      "completions/min_terminated_length": 36.0,
+      "epoch": 0.0007460141529542161,
+      "frac_reward_zero_std": 0.75,
+      "grad_norm": 0.018799597397446632,
+      "kl": 0.0,
+      "learning_rate": 2.314506460636739e-07,
+      "loss": 0.006735410541296005,
+      "num_tokens": 139107.0,
+      "reward": -1.4901161193847656e-08,
+      "reward_std": 0.3785453140735626,
+      "rewards/TranscriptCorrectionGrpoReward/mean": -1.4901161193847656e-08,
+      "rewards/TranscriptCorrectionGrpoReward/std": 1.0160009860992432,
+      "step": 140
+    },
+    {
+      "eval_cer_subset": 0.01283241324278991,
+      "eval_cer_subset_edit_distance": 788,
+      "eval_cer_subset_groups": 250,
+      "eval_cer_subset_items": 250,
+      "eval_cer_subset_ref_chars": 61407,
+      "eval_loss": 0.039553486458817955,
+      "step": 141
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 750656,
+  "num_input_tokens_seen": 139107,
+  "num_train_epochs": 4,
+  "save_steps": 141,
+  "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": 2,
+  "trial_name": null,
+  "trial_params": null
+}

meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37045cadb393fa098f605486af145c8b02ad4b102f7c97ed0190b06ee6961699
+size 6737

meta-llama__llama-3.2-1b/grpo/run_manifest.json

@@ -1,31 +1,25 @@
 {
   "schema_version": 1,
-  "run_id": "525fad5f-fe8b-4a6f-80c1-1f0dd26b7111",
+  "run_id": "ae290391-d91e-4cff-b1c0-a7e1df5cabe9",
   "run_type": "grpo",
   "profile_name": "llama3_2_1b_base_grpo",
   "model_name": "meta-llama/Llama-3.2-1B",
   "status": "running",
   "best_metric_name": "eval_cer_subset",
-  "best_metric_value": 0.012816128454410734,
-  "best_step": 24,
-  "best_checkpoint_step": 24,
-  "latest_checkpoint_step": 24,
+  "best_metric_value": 0.01283241324278991,
+  "best_step": 141,
+  "best_checkpoint_step": 141,
+  "latest_checkpoint_step": 141,
   "checkpoints": [
     {
-      "step": 12,
-      "epoch": 6.394202589652049e-05,
-      "repo_path": "meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-12",
-      "timestamp": "2026-04-08T19:27:43.963032+00:00"
-    },
-    {
-      "step": 24,
-      "epoch": 0.00012788405179304098,
-      "repo_path": "meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-24",
-      "timestamp": "2026-04-08T19:29:00.076745+00:00"
+      "step": 141,
+      "epoch": 0.0007513428254753176,
+      "repo_path": "meta-llama__llama-3.2-1b/grpo/checkpoints/checkpoint-141",
+      "timestamp": "2026-04-08T20:12:31.968726+00:00"
     }
   ],
   "epoch_artifacts": [],
   "resume_history": [],
-  "created_at": "2026-04-08T19:26:15.718549+00:00",
-  "updated_at": "2026-04-08T19:29:00.076759+00:00"
+  "created_at": "2026-04-08T20:07:51.612775+00:00",
+  "updated_at": "2026-04-08T20:12:31.968742+00:00"
 }