Downloads · 30 days
131
48% of all-time downloads
ruadapt_qwen2.5_3B_bpe_32000_full_lr3e4_bs256
RefalMachine/ruadapt_qwen2.5_3B_bpe_32000_full_lr3e4_bs256
ruadapt_qwen2.5_3B_bpe_32000_full_lr3e4_bs256 is a machine learning model from RefalMachine. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
should probably proofread and complete it, then remove this comment. --
All-time downloads
275
Public
Parameters
2.8B
11.5 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors11.4 GB · 93%
At a glance
- Model type
- qwen2
- Access
- Public
- Created
- Oct 10, 2024
- Updated
- Oct 10, 2024
- SHA
- 7880b17b
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
ruadapt_qwen2.5_3B_bpe_32000_full_lr3e4_bs256
This model is a fine-tuned version of RefalMachine/ruadapt_qwen2.5_3B_bpe_32000_mean_init on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.3379
- Accuracy: 0.5208
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: 0.0003
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 64
- gradient_accumulation_steps: 2
- total_train_batch_size: 256
- total_eval_batch_size: 128
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 1.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 0.0 | 1 | 5.9470 | 0.1656 |
| 2.5014 | 0.16 | 2000 | 2.3753 | 0.5161 |
| 2.4916 | 0.32 | 4000 | 2.3500 | 0.5191 |
| 2.4589 | 0.48 | 6000 | 2.3421 | 0.5202 |
| 2.4693 | 0.65 | 8000 | 2.3390 | 0.5208 |
| 2.4574 | 0.81 | 10000 | 2.3380 | 0.5208 |
| 2.4532 | 0.97 | 12000 | 2.3379 | 0.5208 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.3.0a0+6ddf5cf85e.nv24.04
- Datasets 2.18.0
- Tokenizers 0.15.2
- Type
- qwen2
- Created
- Oct 10, 2024
- Updated
- Oct 10, 2024