Downloads · 30 days
3
17% of all-time downloads
ahmedelgebaly/llama-3.1-8b-HotpotQa_SciQ_squad_e1_v2
llama-3.1-8b-HotpotQa_SciQ_squad_e1_v2 is a machine learning model from ahmedelgebaly. 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 llama3.
should probably proofread and complete it, then remove this comment. --
Downloads · 30 days
3
17% of all-time downloads
All-time downloads
18
Public
Repo size
689 MB
Likes
0
Public
Click a slice to open those files.
.bin336 MB · 49%
From the Hugging Face model README
axolotl version: 0.4.1
base_model: meta-llama/Meta-Llama-3.1-8B
lora_model_dir: ahmedelgebaly/llama-3.1-8b-squadv2_SciQ_E1_V4
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
load_in_8bit: false
load_in_4bit: true
strict: false
datasets:
- path: ahmedelgebaly/HotPotQA_Alpaca
type: alpaca
split: train
percentage: 0.7
- path: ahmedelgebaly/SciQ_Alpaca
type: alpaca
split: train
percentage: 0.2
- path: ahmedelgebaly/SQuad_2_Alpaca
type: alpaca
split: train
percentage: 0.1
test_datasets:
- path: ahmedelgebaly/HotPotQA_Alpaca
type: alpaca
split: validation
- path: ahmedelgebaly/SciQ_Alpaca
type: alpaca
split: validation
- path: ahmedelgebaly/SQuad_2_Alpaca
type: alpaca
split: validation
dataset_prepared_path:
output_dir: ./outputs/qlora-out
adapter: qlora
sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true
lora_r: 32
lora_alpha: 64 #Before it was 16
lora_dropout: 0.05
lora_target_modules: #Before it was empty
- q_proj
- k_proj
- v_proj
- o_proj
- gate_proj
- up_proj
- down_proj
lora_target_linear: true
lora_fan_in_fan_out:
wandb_project: llama-3.1-8b-HotpotQa_SciQ_squad_e1_v2
wandb_entity:
wandb_watch:
wandb_name: llama-3.1-8b-HotpotQa_SciQ_squad_e1_v2
wandb_log_model:
hub_model_id: ahmedelgebaly/llama-3.1-8b-HotpotQa_SciQ_squad_e1_v2
gradient_accumulation_steps: 4
micro_batch_size: 4
num_epochs: 1
optimizer: paged_adamw_32bit
lr_scheduler: cosine
learning_rate: 0.0002
train_on_inputs: false
group_by_length: true #Before it was false
bf16: auto
tf32: false
gradient_checkpointing: true
flash_attention: true
warmup_steps: 50 #Before it was 10
evals_per_epoch: 4
saves_per_epoch: 1
weight_decay: 0.0
special_tokens:
pad_token: "<|end_of_text|>"
</details><br>
This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B on the None 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 |
|---|---|---|---|
| No log | 0.0004 | 1 | 1.2136 |
| 0.7524 | 0.2503 | 571 | 0.8545 |
| 0.6904 | 0.5006 | 1142 | 0.8281 |
| 0.6825 | 0.7509 | 1713 | 0.8072 |