Merge pull request #435 from VainF/v2.0

Add SliceOp; Add Phi-3 & Qwen-2

examples/LLMs/prune_llama.py → examples/LLMs/prune_llm.py

@@ -247,14 +247,15 @@ def pattern_match(patterns, source_list):
 print('accelerate', version('accelerate'))
 print('# of gpus: ', torch.cuda.device_count())
 
-def get_llm(model_name):
+def get_llm(model_name, max_seq_len=None):
     model = AutoModelForCausalLM.from_pretrained(
         model_name, 
         torch_dtype=torch.float16, 
         device_map="auto"
     )
 
-    model.seqlen = model.config.max_position_embeddings 
+    model.seqlen = max(max_seq_len, model.config.max_position_embeddings) if max_seq_len is not None else model.config.max_position_embeddings
+     # avoid OOM, feel free to change this
     return model
 
 def main():
@@ -266,6 +267,7 @@ def main():
     parser.add_argument('--save', type=str, default=None, help='Path to save results.')
     parser.add_argument('--save_model', type=str, default=None, help='Path to save the pruned model.')
     parser.add_argument("--eval_zero_shot", action="store_true")
+    parser.add_argument("--max_seq_len", type=int, default=None)
     args = parser.parse_args()
 
     # Setting seeds for reproducibility
@@ -274,7 +276,7 @@ def main():
 
     model_name = args.model.split("/")[-1]
     print(f"loading llm model {args.model}")
-    model = get_llm(args.model)       
+    model = get_llm(args.model, max_seq_len=args.max_seq_len)       
     model.eval()
     tokenizer = AutoTokenizer.from_pretrained(args.model, use_fast=False)
     device = torch.device("cuda:0")
@@ -291,11 +293,22 @@ def main():
     inputs = torch.tensor(tokenizer.encode(text)).unsqueeze(0).to(model.device)
     import torch_pruning as tp 
     num_heads = {}
+    out_channel_groups = {}
+    seperate_qkv = False
     for name, m in model.named_modules():
         if name.endswith("self_attn"):
-            num_heads[m.q_proj] = model.config.num_attention_heads
-            num_heads[m.k_proj] = model.config.num_key_value_heads
-            num_heads[m.v_proj] = model.config.num_key_value_heads
+            if hasattr(m, "q_proj"):
+                seperate_qkv = True
+                num_heads[m.q_proj] = model.config.num_attention_heads
+                num_heads[m.k_proj] = model.config.num_key_value_heads
+                num_heads[m.v_proj] = model.config.num_key_value_heads
+            elif hasattr(m, "qkv_proj"):
+                seperate_qkv = False
+                num_heads[m.qkv_proj] = model.config.num_attention_heads
+        if name.endswith('mlp'):
+            if hasattr(m, "gate_up_proj"):
+                out_channel_groups[m.gate_up_proj] = 2
+
     _is_gqa = model.config.num_attention_heads != model.config.num_key_value_heads
     head_pruning_ratio = args.pruning_ratio
     hidden_size_pruning_ratio = args.pruning_ratio
@@ -311,33 +324,56 @@ def main():
         prune_num_heads=True,
         prune_head_dims=False, # we do not prune head dims so that we don't need to prune the ROPE
         head_pruning_ratio=head_pruning_ratio,
+        out_channel_groups=out_channel_groups,
+        round_to=4,
     )
+
+
     #with torch.no_grad():
     #    with importance.compute_importance(model):
     #        calibration_data = "We recommend at least a 1TB hard drive for 4 channels, more if you plan on using 8MP \/ 4K cameras.\nDahua's Lite Series network video recorders offer excellent performance and high recording quality for IP video surveillance applications. For applications where details are critical for identification, this professional NVR provides a powerful processor with up to 4K resolution. Additionally, the NVR features a mouse shortcut operation menu, remote management and control, center storage, edge storage, and back up storage."
     #        calibration_data = torch.tensor(tokenizer.encode(text)).unsqueeze(0).to(model.device)
     #        _ = model(calibration_data)
-    pruner.step()
+
+    #group = pruner.DG.get_pruning_group(model.model.layers[31].mlp.gate_up_proj, tp.prune_linear_out_channels, idxs=list(range(16384)))
+    #print(group)
+
+    for g in pruner.step(interactive=True):
+        print(g)
+        g.prune()
 
     # Update model attributes    
     model.config.hidden_size = model.lm_head.in_features
     for name, m in model.named_modules():
         if name.endswith("self_attn"):
-            m.hidden_size = m.q_proj.out_features
+            if seperate_qkv:
+                m.hidden_size = m.q_proj.out_features
+            else:
+                m.hidden_size = m.qkv_proj.out_features // 3        
             m.num_heads = m.hidden_size // m.head_dim
             model.config.num_attention_heads = m.num_heads
             #m.head_dim = m.q_proj.out_features // m.num_heads
             if not _is_gqa:
                 m.num_key_value_heads = m.num_heads
             m.num_key_value_groups = m.num_heads // m.num_key_value_heads
         elif name.endswith("mlp"):
-            model.config.intermediate_size = m.gate_proj.out_features
+            if hasattr(m, "gate_proj"):
+                m.hidden_size = m.gate_proj.out_features
+            elif hasattr(m, "gate_up_proj"):
+                m.hidden_size = m.gate_up_proj.in_features
+            else:
+                raise ValueError("Unknown mlp layer")
+
     if not _is_gqa:
         model.config.num_key_value_heads = model.config.num_attention_heads
     print("----------------- After Pruning -----------------")
     print(model)
     print(model.config)
 
+
+    del pruner
+    torch.cuda.empty_cache()
+    model.eval()
     num_params = sum(p.numel() for p in model.parameters())
     print(f"num_params {num_params}")
     ppl_test = eval_ppl(args, model, tokenizer, device)

examples/LLMs/readme.md

@@ -1,7 +1,16 @@
-# Prune Large Language Models
+# Pruning Large Language Models
 
 This example provides a minimal example of pruning large language models with magnitude-based pruning. We use the `transformers` library to load the model and the `datasets` library to evaluate the Perplexity with `Wikitext2`. **For more comprehensive examples of Gradient-based pruning or finetuning, please refer to [LLM-Pruner](https://github.com/horseee/LLM-Pruner)**.
 
+This script has been tested with the following models:
+
+1. [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B)
+2. [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf)
+3. [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct)
+4. [Qwen/Qwen2-7B](https://huggingface.co/Qwen/Qwen2-7B)
+
+
+
 ## 0. Requirements
 
 ```bash
@@ -13,7 +22,7 @@ pip install transformers datasets
 ### Llama-3 8B
 
 ```bash
-python prune_llama.py --model meta-llama/Meta-Llama-3-8B --pruning_ratio 0.5
+python prune_llm.py --model meta-llama/Meta-Llama-3-8B --pruning_ratio 0.5
 ```
 
 <details>
@@ -120,7 +129,7 @@ wikitext perplexity 552648.25
 ### Llama-2 7B
 
 ```bash
-python prune_llama.py --model meta-llama/Llama-2-7b-hf --pruning_ratio 0.5
+python prune_llm.py --model meta-llama/Llama-2-7b-hf --pruning_ratio 0.5
 ```
 
 
@@ -224,3 +233,219 @@ wikitext perplexity 8479.0673828125
 </details>
 
 
+### microsoft/Phi-3-mini-4k-instruct
+
+```bash
+python prune_llm.py --model microsoft/Phi-3-mini-4k-instruct --pruning_ratio 0.5
+```
+
+
+<details>
+<summary>Output:</summary>
+
+```
+----------------- Before Pruning -----------------
+Phi3ForCausalLM(
+  (model): Phi3Model(
+    (embed_tokens): Embedding(32064, 3072, padding_idx=32000)
+    (embed_dropout): Dropout(p=0.0, inplace=False)
+    (layers): ModuleList(
+      (0-31): 32 x Phi3DecoderLayer(
+        (self_attn): Phi3Attention(
+          (o_proj): Linear(in_features=3072, out_features=3072, bias=False)
+          (qkv_proj): Linear(in_features=3072, out_features=9216, bias=False)
+          (rotary_emb): Phi3RotaryEmbedding()
+        )
+        (mlp): Phi3MLP(
+          (gate_up_proj): Linear(in_features=3072, out_features=16384, bias=False)
+          (down_proj): Linear(in_features=8192, out_features=3072, bias=False)
+          (activation_fn): SiLU()
+        )
+        (input_layernorm): Phi3RMSNorm()
+        (resid_attn_dropout): Dropout(p=0.0, inplace=False)
+        (resid_mlp_dropout): Dropout(p=0.0, inplace=False)
+        (post_attention_layernorm): Phi3RMSNorm()
+      )
+    )
+    (norm): Phi3RMSNorm()
+  )
+  (lm_head): Linear(in_features=3072, out_features=32064, bias=False)
+)
+----------------- After Pruning -----------------
+Token indices sequence length is longer than the specified maximum sequence length for this model (2824490 > 4096). Running this sequence through the model will result in indexing errors
+Phi3ForCausalLM(
+  (model): Phi3Model(
+    (embed_tokens): Embedding(32064, 1536, padding_idx=32000)
+    (embed_dropout): Dropout(p=0.0, inplace=False)
+    (layers): ModuleList(
+      (0-31): 32 x Phi3DecoderLayer(
+        (self_attn): Phi3Attention(
+          (o_proj): Linear(in_features=1536, out_features=1536, bias=False)
+          (qkv_proj): Linear(in_features=1536, out_features=4608, bias=False)
+          (rotary_emb): Phi3RotaryEmbedding()
+        )
+        (mlp): Phi3MLP(
+          (gate_up_proj): Linear(in_features=1536, out_features=8192, bias=False)
+          (down_proj): Linear(in_features=4096, out_features=1536, bias=False)
+          (activation_fn): SiLU()
+        )
+        (input_layernorm): Phi3RMSNorm()
+        (resid_attn_dropout): Dropout(p=0.0, inplace=False)
+        (resid_mlp_dropout): Dropout(p=0.0, inplace=False)
+        (post_attention_layernorm): Phi3RMSNorm()
+      )
+    )
+    (norm): Phi3RMSNorm()
+  )
+  (lm_head): Linear(in_features=1536, out_features=32064, bias=False)
+)
+Phi3Config {
+  "_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
+  "architectures": [
+    "Phi3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "auto_map": {
+    "AutoConfig": "microsoft/Phi-3-mini-4k-instruct--configuration_phi3.Phi3Config",
+    "AutoModelForCausalLM": "microsoft/Phi-3-mini-4k-instruct--modeling_phi3.Phi3ForCausalLM"
+  },
+  "bos_token_id": 1,
+  "embd_pdrop": 0.0,
+  "eos_token_id": 32000,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 4096,
+  "model_type": "phi3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 16,
+  "original_max_position_embeddings": 4096,
+  "pad_token_id": 32000,
+  "resid_pdrop": 0.0,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "sliding_window": 2047,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.36.2",
+  "use_cache": true,
+  "vocab_size": 32064
+}
+
+num_params 1004570112
+evaluating on wikitext2
+nsamples 83
+sample 0
+sample 50
+wikitext perplexity 92795.3984375
+```
+
+</details>
+
+### Qwen/Qwen2-7B
+
+```bash
+python prune_llm.py --model Qwen/Qwen2-7B --pruning_ratio 0.5
+```
+
+
+<details>
+<summary>Output:</summary>
+
+```
+----------------- Before Pruning -----------------
+Qwen2ForCausalLM(
+  (model): Qwen2Model(
+    (embed_tokens): Embedding(152064, 3584)
+    (layers): ModuleList(
+      (0-27): 28 x Qwen2DecoderLayer(
+        (self_attn): Qwen2SdpaAttention(
+          (q_proj): Linear(in_features=3584, out_features=3584, bias=True)
+          (k_proj): Linear(in_features=3584, out_features=512, bias=True)
+          (v_proj): Linear(in_features=3584, out_features=512, bias=True)
+          (o_proj): Linear(in_features=3584, out_features=3584, bias=False)
+          (rotary_emb): Qwen2RotaryEmbedding()
+        )
+        (mlp): Qwen2MLP(
+          (gate_proj): Linear(in_features=3584, out_features=18944, bias=False)
+          (up_proj): Linear(in_features=3584, out_features=18944, bias=False)
+          (down_proj): Linear(in_features=18944, out_features=3584, bias=False)
+          (act_fn): SiLU()
+        )
+        (input_layernorm): Qwen2RMSNorm((3584,), eps=1e-06)
+        (post_attention_layernorm): Qwen2RMSNorm((3584,), eps=1e-06)
+      )
+    )
+    (norm): Qwen2RMSNorm((3584,), eps=1e-06)
+    (rotary_emb): Qwen2RotaryEmbedding()
+  )
+  (lm_head): Linear(in_features=3584, out_features=152064, bias=False)
+)
+----------------- After Pruning -----------------
+Qwen2ForCausalLM(
+  (model): Qwen2Model(
+    (embed_tokens): Embedding(152064, 1792)
+    (layers): ModuleList(
+      (0-27): 28 x Qwen2DecoderLayer(
+        (self_attn): Qwen2SdpaAttention(
+          (q_proj): Linear(in_features=1792, out_features=2048, bias=True)
+          (k_proj): Linear(in_features=1792, out_features=512, bias=True)
+          (v_proj): Linear(in_features=1792, out_features=512, bias=True)
+          (o_proj): Linear(in_features=2048, out_features=1792, bias=False)
+          (rotary_emb): Qwen2RotaryEmbedding()
+        )
+        (mlp): Qwen2MLP(
+          (gate_proj): Linear(in_features=1792, out_features=9472, bias=False)
+          (up_proj): Linear(in_features=1792, out_features=9472, bias=False)
+          (down_proj): Linear(in_features=9472, out_features=1792, bias=False)
+          (act_fn): SiLU()
+        )
+        (input_layernorm): Qwen2RMSNorm((1792,), eps=1e-06)
+        (post_attention_layernorm): Qwen2RMSNorm((1792,), eps=1e-06)
+      )
+    )
+    (norm): Qwen2RMSNorm((1792,), eps=1e-06)
+    (rotary_emb): Qwen2RotaryEmbedding()
+  )
+  (lm_head): Linear(in_features=1792, out_features=152064, bias=False)
+)
+Qwen2Config {
+  "_attn_implementation_autoset": true,
+  "_name_or_path": "Qwen/Qwen2-7B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1792,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.46.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}
+
+num_params 2227887872
+```
+
+</details>
+

examples/torchvision_models/torchvision_global_pruning.py

@@ -16,8 +16,6 @@
 )
 from torchvision.models.detection.fcos import fcos_resnet50_fpn
 from torchvision.models.detection.keypoint_rcnn import keypointrcnn_resnet50_fpn
-from torchvision.models.detection.mask_rcnn import maskrcnn_resnet50_fpn_v2
-from torchvision.models.detection.retinanet import retinanet_resnet50_fpn_v2
 from torchvision.models.alexnet import alexnet
 
 from torchvision.models.vision_transformer import (

examples/torchvision_models/torchvision_pruning.py

@@ -16,8 +16,6 @@
 )
 from torchvision.models.detection.fcos import fcos_resnet50_fpn
 from torchvision.models.detection.keypoint_rcnn import keypointrcnn_resnet50_fpn
-from torchvision.models.detection.mask_rcnn import maskrcnn_resnet50_fpn_v2
-from torchvision.models.detection.retinanet import retinanet_resnet50_fpn_v2
 from torchvision.models.alexnet import alexnet
 
 from torchvision.models.vision_transformer import (