cot-transduction-only-arc
This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on the barc0/trans_only_cot_100k-gpt4omini-description, the barc0/trans_only_cot_100k-gpt4-description and the barc0/trans_only_cot_200k_HEAVY_gpt4o-description datasets. It achieves the following results on the evaluation set:
- Loss: 0.0253
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.0308 | 0.9998 | 2994 | 0.0342 |
| 0.0276 | 2.0 | 5989 | 0.0253 |
| 0.0116 | 2.9995 | 8982 | 0.0253 |
Framework versions
- Transformers 4.45.0.dev0
- Pytorch 2.4.0+cu121
- Datasets 3.2.0
- Tokenizers 0.19.1
- Downloads last month
- 6
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for barc0/cot-transduction-only-arc
Base model
meta-llama/Llama-3.1-8B
Finetuned
meta-llama/Llama-3.1-8B-Instruct