Create README.md

README.md

@@ -0,0 +1,42 @@
+---
+tags:
+- gptq
+language:
+- en
+base_model: Sao10K/L3-8B-Stheno-v3.2
+---
+
+Original Model: https://huggingface.co/Sao10K/L3-8B-Stheno-v3.2
+
+Quantized with AutoGPTQ 128g wikitext2, using the script from https://aphrodite.pygmalion.chat/pages/quantization/quantization-methods.html#gptq
+
+Script:
+```python
+from datasets import load_dataset
+from transformers import AutoTokenizer
+
+from auto_fp8 import AutoFP8ForCausalLM, BaseQuantizeConfig
+
+pretrained_model_dir = "Sao10K/L3-8B-Stheno-v3.2"
+quantized_model_dir = "L3-8B-Stheno-v3.2-FP8"
+
+tokenizer = AutoTokenizer.from_pretrained(pretrained_model_dir, use_fast=True, model_max_length=4096)
+tokenizer.pad_token = tokenizer.eos_token
+
+ds = load_dataset("mgoin/ultrachat_2k", split="train_sft").select(range(512))
+examples = [tokenizer.apply_chat_template(batch["messages"], tokenize=False) for batch in ds]
+examples = tokenizer(examples, padding=True, truncation=True, return_tensors="pt").to("cuda")
+
+quantize_config = BaseQuantizeConfig(
+ quant_method="fp8",
+ activation_scheme="static",
+ ignore_patterns=["re:.*lm_head"],
+)
+
+model = AutoFP8ForCausalLM.from_pretrained(
+ pretrained_model_dir, quantize_config=quantize_config
+)
+
+model.quantize(examples)
+model.save_quantized(quantized_model_dir)
+```