YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
Quantization made by Richard Erkhov.
Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0 - GGUF
- Model creator: https://huggingface.co/Na0s/
- Original model: https://huggingface.co/Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0/
Original model description:
library_name: transformers datasets: - teknium/openhermes pipeline_tag: text-generation license: apache-2.0 base_model: Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-2.0
Model Card for Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0:
Model Details:
Model Description:
- Finetuned from model: Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-2.0 on teknium/openhermes.
- We pruned the 4 layers of meta-llama/Meta-Llama-3.1-8B that had the less impact on the performance of the model according to the paper The Unreasonable Ineffectiveness of the Deeper Layers.
- We have therefore 1.09B parameters less than the foundation model, which means less memory needed, faster training and less latency during inference mode.
- We then recovered the performance loss induced by the pruning process by fine-tuning (from 0.2642 MMLU-Pro 0-shot to 0.3120), this step is called healing the pruned model.
Upcoming Work:
- More healing through SFT/DPO/TPO to see if we can get closer to the meta-llama/Meta-Llama-3.1-8B performance (which has an MMLU-Pro 0-shot of 0.3659). (In Progress)
- Evaluate on benchmarks other than MMLU-PRO 0-shot (Unfortunately lighteval is broken right now issue #191, issue #213).
- Compare the same exact process when applied to meta-llama/LLama-3.1-70B.
Training Details:
model = FastLanguageModel.get_peft_model(
model,
r = 4,
target_modules = ["q_proj", "k_proj", "v_proj", "o_proj",
"gate_proj", "up_proj", "down_proj",],
lora_alpha = 4,
lora_dropout = 0.05,
bias = "none",
use_gradient_checkpointing = "unsloth",
random_state = 3407,
use_rslora = False,
loftq_config = None,
)
from trl import SFTTrainer
from transformers import TrainingArguments
from unsloth import is_bfloat16_supported
trainer = SFTTrainer(
model = model,
tokenizer = tokenizer,
train_dataset = dataset,
dataset_text_field = "completion",
max_seq_length = max_seq_length,
dataset_num_proc = 2,
packing = False,
args = TrainingArguments(
per_device_train_batch_size = 10,
gradient_accumulation_steps = 4,
warmup_steps = 5,
max_steps=5000,
learning_rate = 2e-4,
fp16 = not is_bfloat16_supported(),
bf16 = is_bfloat16_supported(),
logging_steps = 1,
optim = "adamw_8bit",
weight_decay = 0.01,
lr_scheduler_type = "cosine",
seed = 3407,
output_dir = "outputs_4",
push_to_hub=True,
hub_always_push=True,
),
)
Training Data:
Memory and Latency gain (Using Optimum-Benchmark):
Load Mode Memory Metrics
| Model | Max Global VRAM (MB) | Max Process VRAM (MB) | Max Reserved VRAM (MB) | Max Allocated VRAM (MB) |
|---|---|---|---|---|
| Llama-3.1-8B | 18521.98 | 16630.42 | 16196.30 | 16060.54 |
| Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0 | 16319.97 | 14428.41 | 13994.30 | 13879.42 |
Inference Mode Latency Metrics
| Model | Latency Mean (s) | Throughput (tokens/s) |
|---|---|---|
| Llama-3.1-8B | 0.8104 | 38.2536 |
| Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0 | 0.5530 | 56.0570 |
Evaluation:
- (Foundation model) MMLU Pro 0-shot of meta-llama/Meta-Llama-3.1-8B: 0.3659
- (Pruned model) MMLU Pro 0-shot of Na0s/Llama-3.1-8B-Pruned-4-Layers: 0.2642
- (Healed model) MMLU Pro 0-shot of Na0s/Llama-3.1-8B-Pruned-4-Layers_LoRA-PEFT-3.0: 0.3120
Evaluation Data and Process:
- TIGER-AI-Lab/MMLU-Pro.
- More classic benchmarks like ARC incoming.
Environmental Impact:
Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
- Downloads last month
- 240