Asynchronous RLHF
Collection
Models and datasets for asynchronous rlhf paper, see code at https://github.com/mnoukhov/async_rlhf
•
10 items
•
Updated
pythia1b-rm-tldr6.9b
This model is a fine-tuned version of mnoukhov/pythia1b-sft-tldr on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4685
- Accuracy: 0.7693
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: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5365 | 0.2006 | 291 | 0.5420 | 0.7271 |
| 0.4521 | 0.4011 | 582 | 0.5034 | 0.7485 |
| 0.3994 | 0.6017 | 873 | 0.4893 | 0.7577 |
| 0.3596 | 0.8022 | 1164 | 0.4685 | 0.7693 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
- Downloads last month
- 14
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for mnoukhov/pythia1b-rm-tldr6.9b
Base model
EleutherAI/pythia-1b-deduped
Finetuned
mnoukhov/pythia1b-sft-tldr