split

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

Author: kylesayrs

README.md

@@ -86,7 +86,8 @@ Applying quantization with `llmcompressor`:
 * [Weight only quantization to `int4` using AWQ](examples/awq/README.md)
 * [Weight only quantization to `int4` using AutoRound](examples/autoround/quantization_w4a16/README.md)
 * [KV Cache quantization to `fp8`](examples/quantization_kv_cache/README.md)
-* [Attention quantization to `fp8` (experimental)](examples/quantization_attention/README.md)
+* [KV Cache quantization to `fp8` using per-head](examples/quantization_kv_cache/llama3_fp8_head_kv_example.py)
+* [Attention quantization to `fp8`](examples/quantization_attention/README.md)
 * [Attention quantization to `nvfp4` with SpinQuant (experimental)](experimental/attention/README.md)
 * [Quantizing MoE LLMs](examples/quantizing_moe/README.md)
 * [Quantizing Vision-Language Models](examples/multimodal_vision/README.md)

examples/quantization_attention/llama3_attention.py

@@ -56,7 +56,7 @@ def tokenize(sample):
         "attention": QuantizationScheme(
             targets=["LlamaAttention"],
             input_activations=QuantizationArgs(
-                num_bits=8, type="float", strategy="attn_head"
+                num_bits=8, type="float", strategy="tensor"
             ),
         )
     }

examples/quantization_kv_cache/llama3_fp8_head_kv_example.py

@@ -0,0 +1,83 @@
+from compressed_tensors.offload import dispatch_model
+from compressed_tensors.quantization import QuantizationArgs
+from datasets import load_dataset
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+from llmcompressor import oneshot
+from llmcompressor.modifiers.quantization import QuantizationModifier
+
+# Select model and load it.
+model_id = "meta-llama/Meta-Llama-3-8B-Instruct"
+model = AutoModelForCausalLM.from_pretrained(model_id, dtype="auto")
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+# Select calibration dataset.
+DATASET_ID = "HuggingFaceH4/ultrachat_200k"
+DATASET_SPLIT = "train_sft"
+
+# Select number of samples. 512 samples is a good place to start.
+# Increasing the number of samples can improve accuracy.
+NUM_CALIBRATION_SAMPLES = 512
+MAX_SEQUENCE_LENGTH = 2048
+
+# Load dataset and preprocess.
+ds = load_dataset(DATASET_ID, split=f"{DATASET_SPLIT}[:{NUM_CALIBRATION_SAMPLES}]")
+ds = ds.shuffle(seed=42)
+
+
+def preprocess(example):
+    return {
+        "text": tokenizer.apply_chat_template(
+            example["messages"],
+            tokenize=False,
+        )
+    }
+
+
+ds = ds.map(preprocess)
+
+
+# Tokenize inputs.
+def tokenize(sample):
+    return tokenizer(
+        sample["text"],
+        padding=False,
+        max_length=MAX_SEQUENCE_LENGTH,
+        truncation=True,
+        add_special_tokens=False,
+    )
+
+
+ds = ds.map(tokenize, remove_columns=ds.column_names)
+
+# Configure the quantization algorithm to run.
+recipe = QuantizationModifier(
+    targets="Linear",
+    scheme="FP8_DYNAMIC",
+    ignore=["lm_head"],
+    kv_cache_scheme=QuantizationArgs(num_bits=8, type="float", strategy="attn_head"),
+)
+
+# Apply algorithms.
+oneshot(
+    model=model,
+    dataset=ds,
+    recipe=recipe,
+    max_seq_length=MAX_SEQUENCE_LENGTH,
+    num_calibration_samples=NUM_CALIBRATION_SAMPLES,
+)
+
+# Confirm generations of the quantized model look sane.
+print("\n\n")
+print("========== SAMPLE GENERATION ==============")
+dispatch_model(model)
+sample = tokenizer("Hello my name is", return_tensors="pt")
+sample = {key: value.to(model.device) for key, value in sample.items()}
+output = model.generate(**sample, max_new_tokens=100)
+print(tokenizer.decode(output[0]))
+print("==========================================\n\n")
+
+# Save to disk compressed.
+SAVE_DIR = model_id.rstrip("/").split("/")[-1] + "-attention-fp8-head"
+model.save_pretrained(SAVE_DIR, save_compressed=True)
+tokenizer.save_pretrained(SAVE_DIR)