---
library_name: transformers
tags:
- generated_from_trainer
datasets:
- 2025-01_conversations_truncated.jsonl
model-index:
- name: outputs/
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>

axolotl version: `0.6.0`
```yaml
base_model: ./meta-llama_Llama-3.2-3B
# optionally might have model_type or tokenizer_type
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer
# Automatically upload checkpoint and final model to HF
# hub_model_id: username/custom_model_name

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: 2025-01_conversations_truncated.jsonl
    type: chat_template
    chat_template: llama3
    field_messages: conversations
    message_field_role: from
    message_field_content: value
    roles:
      user:
        - human
      assistant:
        - gpt
      system:
        - system
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./outputs/
dataset_prepared_path: last_run_prepared

sequence_len: 4096
eval_sample_packing: false
sample_packing: true
pad_to_sequence_len: true

wandb_project: JVCGPT Light 3b base
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 4
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.000007

train_on_inputs: true
group_by_length: false
bf16: auto
fp16:
tf32: false

gradient_checkpointing: unsloth
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
s2_attention:

warmup_steps: 100
eval_table_size:
saves_per_epoch: 2
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
   pad_token: <|end_of_text|>
save_safetensors: true
save_total_limit: 10
```

</details><br>

# outputs/

This model was trained from scratch on the 2025-01_conversations_truncated.jsonl dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1520

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 7e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- total_eval_batch_size: 8
- optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 4

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.6055        | 1.0006 | 789  | 1.1893          |
| 0.5619        | 2.0006 | 1578 | 1.1576          |
| 0.4873        | 3.0006 | 2367 | 1.1522          |
| 1.2133        | 3.9917 | 3148 | 1.1520          |


### Framework versions

- Transformers 4.47.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0