Downloads · 30 days
3
11% of all-time downloads
cgifbribcgfbi/alpha32_r64_lr0.00002_Qwen3-32B-ablit_textbook_5000
alpha32_r64_lr0.00002_Qwen3-32B-ablit_textbook_5000 is a machine learning model from cgifbribcgfbi. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
Downloads · 30 days
3
11% of all-time downloads
All-time downloads
28
Public
Repo size
8.6 GB
Likes
0
Public
Click a slice to open those files.
.safetensors2.1 GB · 99%
From the Hugging Face model README
axolotl version: 0.9.1
base_model: roslein/Qwen3-32B-abliterated
load_in_8bit: false
load_in_4bit: true
adapter: qlora
wandb_name: Qwen3-32B-ablit_outputs_axolotl_ft_alpha32_r64_lr0.00002_Qwen3-32B-ablit_textbook_5000
output_dir: ./outputs/out/Qwen3-32B-ablit_outputs_axolotl_ft_alpha32_r64_lr0.00002_Qwen3-32B-ablit_textbook_5000
hub_model_id: cgifbribcgfbi/alpha32_r64_lr0.00002_Qwen3-32B-ablit_textbook_5000
tokenizer_type: AutoTokenizer
push_dataset_to_hub:
strict: false
datasets:
- path: chemNLP/chemistry-bookshelves-merged
type: completion
dataset_prepared_path: last_run_prepared
val_set_size: 0.04
save_safetensors: true
sequence_len: 2700
sample_packing: true
pad_to_sequence_len: true
lora_r: 64
lora_alpha: 32
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true
wandb_mode:
wandb_project: finetune-sweep
wandb_entity: gpoisjgqetpadsfke
wandb_watch:
wandb_run_id:
wandb_log_model:
gradient_accumulation_steps: 1
micro_batch_size: 4 # This will be automatically adjusted based on available GPU memory
num_epochs: 4
optimizer: adamw_torch_fused
lr_scheduler: cosine
learning_rate: 0.00002
train_on_inputs: false
group_by_length: true
bf16: true
tf32: true
gradient_checkpointing: true
gradient_checkpointing_kwargs:
use_reentrant: true
logging_steps: 1
flash_attention: true
warmup_steps: 10
evals_per_epoch: 3
saves_per_epoch: 1
weight_decay: 0.01
fsdp:
- full_shard
- auto_wrap
fsdp_config:
fsdp_limit_all_gathers: true
fsdp_sync_module_states: true
fsdp_offload_params: false
fsdp_use_orig_params: false
fsdp_cpu_ram_efficient_loading: true
fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
fsdp_transformer_layer_cls_to_wrap: Qwen3DecoderLayer
fsdp_state_dict_type: FULL_STATE_DICT
fsdp_sharding_strategy: FULL_SHARD
special_tokens:
pad_token: <|finetune_right_pad_id|>
</details><br>
This model is a fine-tuned version of roslein/Qwen3-32B-abliterated on the chemNLP/chemistry-bookshelves-merged dataset. It achieves the following results on the evaluation set:
More information needed
More information needed
More information needed
The following hyperparameters were used during training:
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.025 | 0.0032 | 1 | 1.0347 |
| 0.8223 | 0.3354 | 106 | 0.8981 |
| 1.0176 | 0.6709 | 212 | 0.8696 |
| 0.6425 | 1.0063 | 318 | 0.8556 |
| 0.9802 | 1.3418 | 424 | 0.8463 |
| 1.1295 | 1.6772 | 530 | 0.8445 |
| 0.6917 | 2.0127 | 636 | 0.8415 |
| 1.0986 | 2.3481 | 742 | 0.8374 |
| 1.0146 | 2.6835 | 848 | 0.8366 |
| 0.857 | 3.0190 | 954 | 0.8360 |
| 0.9889 | 3.3544 | 1060 | 0.8344 |
| 1.2498 | 3.6899 | 1166 | 0.8344 |