---
library_name: transformers
tags:
- generated_from_trainer
model-index:
- name: lustre/fswork/projects/rech/dgo/udv55np/ift/Qwen3-235B-A22B/Qwen2.5-0.5B/0
  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: `0.12.0.dev0`
```yaml
base_model: /lustre/fswork/projects/rech/qwv/udv55np/Qwen/Qwen2.5-0.5B

datasets:
- path: /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking
  ds_type: json
  type: chat_template
  field_messages: conversations
  data_files:
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0007.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0009.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0005.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0006.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0014.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0010.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0012.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0008.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0001.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0002.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0013.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0015.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0004.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0011.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0000.jsonl
  - /lustre/fswork/projects/rech/qwv/udv55np/dataset/ift/Qwen3-235B-A22B/no_thinking/0003.jsonl

dataset_prepared_path: /lustre/fsn1/projects/rech/dgo/udv55np/dataset/Qwen3-235B-A22B/Qwen2.5-0.5B/mix_0
chat_template: qwen_25

shuffle_merged_datasets: true
output_dir: /lustre/fswork/projects/rech/dgo/udv55np/ift/Qwen3-235B-A22B/Qwen2.5-0.5B/0

sequence_len: 16384
sample_packing: true

gradient_accumulation_steps: 1
micro_batch_size: 1
num_epochs: 1
auto_resume_from_checkpoints: true

optimizer: adamw_torch_fused
lr_scheduler: warmup_stable_decay
learning_rate: 1e-5
lr_scheduler_kwargs:
  num_decay_steps: 300
  min_lr_ratio: 0.1
warmup_steps: 150

bf16: true
tf32: false

gradient_checkpointing: true
logging_steps: 10
flash_attention: true

evals_per_epoch: 0
save_steps: 0.2
save_total_limit: 20
save_only_model: false

use_tensorboard: true
deepspeed: /lustre/fswork/projects/rech/qwv/udv55np/axolotl/zero3.json
special_tokens:
  bos_token: "<|im_start|>"
  eos_token: "<|im_end|>"
  pad_token: "<|endoftext|>"

```

</details><br>

# lustre/fswork/projects/rech/dgo/udv55np/ift/Qwen3-235B-A22B/Qwen2.5-0.5B/0

This model was trained from scratch on the None dataset.

## 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: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 16
- total_train_batch_size: 16
- total_eval_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: warmup_stable_decay
- lr_scheduler_warmup_steps: 150
- training_steps: 2975

### Training results



### Framework versions

- Transformers 4.53.2
- Pytorch 2.6.0+cu124
- Datasets 4.0.0
- Tokenizers 0.21.1