Datasets:
Modalities:
Tabular
Formats:
json
Size:
< 1K
Tags:
base_model:adapter:/workspace/LLaMA-Factory/models/gemma4-12b/gemma-4-12B-it-qat-q4_0-unquantized
llama-factory
lora
transformers
License:
Search is not available for this dataset
epoch float64 | num_input_tokens_seen int64 | total_flos float64 | train_loss float64 | train_runtime float64 | train_samples_per_second float64 | train_steps_per_second float64 |
|---|---|---|---|---|---|---|
3 | 2,165,632 | 144,065,030,122,438,660 | 0.075502 | 7,826.5931 | 0.383 | 0.024 |
gemma4_esg_qlora
This model is a fine-tuned version of /workspace/LLaMA-Factory/models/gemma4-12b/gemma-4-12B-it-qat-q4_0-unquantized on the esg_cot_train 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: 5e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10.0
- num_epochs: 3.0
Training results
Framework versions
- PEFT 0.18.1
- Transformers 5.13.0.dev0
- Pytorch 2.6.0+cu124
- Datasets 4.0.0
- Tokenizers 0.22.2
- Downloads last month
- 261