Intel Gpu support initialization (#4340)

extensions/multimodal/abstract_pipeline.py

@@ -3,6 +3,7 @@ from typing import List, Optional
 
 import torch
 from PIL import Image
+from transformers import is_torch_xpu_available
 
 
 class AbstractMultimodalPipeline(ABC):
@@ -55,7 +56,7 @@ class AbstractMultimodalPipeline(ABC):
 
     def _get_device(self, setting_name: str, params: dict):
         if params[setting_name] is None:
-            return torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+            return torch.device("cuda:0" if torch.cuda.is_available() else "xpu:0" if is_torch_xpu_available() else "cpu")
         return torch.device(params[setting_name])
 
     def _get_dtype(self, setting_name: str, params: dict):

modules/AutoGPTQ_loader.py

@@ -1,5 +1,6 @@
 from pathlib import Path
 
+from accelerate import is_xpu_available
 from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
 
 import modules.shared as shared
@@ -41,7 +42,7 @@ def load_quantized(model_name):
     # Define the params for AutoGPTQForCausalLM.from_quantized
     params = {
         'model_basename': pt_path.stem,
-        'device': "cuda:0" if not shared.args.cpu else "cpu",
+        'device': "xpu:0" if is_xpu_available() else "cuda:0" if not shared.args.cpu else "cpu",
         'use_triton': shared.args.triton,
         'inject_fused_attention': not shared.args.no_inject_fused_attention,
         'inject_fused_mlp': not shared.args.no_inject_fused_mlp,

modules/GPTQ_loader.py

@@ -5,15 +5,15 @@ from pathlib import Path
 import accelerate
 import torch
 import transformers
+from accelerate import is_xpu_available
+from gptq_for_llama import llama_inference_offload
+from gptq_for_llama.modelutils import find_layers
+from gptq_for_llama.quant import make_quant
 from transformers import AutoConfig, AutoModelForCausalLM
 
 import modules.shared as shared
 from modules.logging_colors import logger
 
-from gptq_for_llama import llama_inference_offload
-from gptq_for_llama.modelutils import find_layers
-from gptq_for_llama.quant import make_quant
-
 
 # This function is a replacement for the load_quant function in the
 # GPTQ-for_LLaMa repository. It supports more models and branches.
@@ -144,7 +144,7 @@ def load_quantized(model_name):
         model = load_quant(str(path_to_model), str(pt_path), shared.args.wbits, shared.args.groupsize, kernel_switch_threshold=threshold)
 
         # accelerate offload (doesn't work properly)
-        if shared.args.gpu_memory or torch.cuda.device_count() > 1:
+        if shared.args.gpu_memory or torch.cuda.device_count() > 1 or (is_xpu_available() and torch.xpu.device_count() > 1):
             if shared.args.gpu_memory:
                 memory_map = list(map(lambda x: x.strip(), shared.args.gpu_memory))
                 max_cpu_memory = shared.args.cpu_memory.strip() if shared.args.cpu_memory is not None else '99GiB'
@@ -163,6 +163,9 @@ def load_quantized(model_name):
 
         # No offload
         elif not shared.args.cpu:
+            if is_xpu_available():
+                model = model.to(torch.device("xpu:0"))
+            else:
                 model = model.to(torch.device('cuda:0'))
 
     return model

modules/LoRA.py

@@ -2,6 +2,7 @@ from pathlib import Path
 
 import torch
 from peft import PeftModel
+from transformers import is_torch_xpu_available
 
 import modules.shared as shared
 from modules.logging_colors import logger
@@ -179,6 +180,9 @@ def add_lora_transformers(lora_names):
                 if torch.backends.mps.is_available():
                     device = torch.device('mps')
                     shared.model = shared.model.to(device)
+                elif is_torch_xpu_available():
+                    device = torch.device("xpu:0")
+                    shared.model = shared.model.to(device)
                 else:
                     shared.model = shared.model.cuda()
 

modules/RWKV.py

@@ -9,6 +9,7 @@ from pathlib import Path
 
 import numpy as np
 from tokenizers import Tokenizer
+from transformers import is_torch_xpu_available
 
 import modules.shared as shared
 from modules.callbacks import Iteratorize
@@ -27,7 +28,7 @@ class RWKVModel:
         pass
 
     @classmethod
-    def from_pretrained(self, path, dtype="fp16", device="cuda"):
+    def from_pretrained(self, path, dtype="bf16" if is_torch_xpu_available() else "fp16", device="xpu" if is_torch_xpu_available() else "cuda"):
         tokenizer_path = Path(f"{path.parent}/20B_tokenizer.json")
         if shared.args.rwkv_strategy is None:
             model = RWKV(model=str(path), strategy=f'{device} {dtype}')

modules/callbacks.py

@@ -5,6 +5,7 @@ from threading import Thread
 
 import torch
 import transformers
+from transformers import is_torch_xpu_available
 
 import modules.shared as shared
 
@@ -92,4 +93,7 @@ class Iteratorize:
 def clear_torch_cache():
     gc.collect()
     if not shared.args.cpu:
+        if is_torch_xpu_available():
+            torch.xpu.empty_cache()
+        else:
             torch.cuda.empty_cache()

modules/logits.py

@@ -1,4 +1,5 @@
 import torch
+from transformers import is_torch_xpu_available
 
 from modules import sampler_hijack, shared
 from modules.logging_colors import logger
@@ -32,11 +33,17 @@ def get_next_logits(prompt, state, use_samplers, previous):
         scores = sampler_hijack.global_scores[-1]
     else:
         if is_non_hf_exllamav2 or is_non_hf_exllamav1:
+            if is_torch_xpu_available():
+                tokens = shared.tokenizer.encode(prompt).to("xpu:0")
+            else:
                 tokens = shared.tokenizer.encode(prompt).cuda()
             scores = shared.model.get_logits(tokens)[-1][-1]
         elif is_non_hf_llamacpp:
             tokens = shared.tokenizer.encode(prompt)
             scores = shared.model.get_logits(tokens)[-1][-1]
+        else:
+            if is_torch_xpu_available():
+                tokens = shared.tokenizer.encode(prompt, return_tensors='pt').to("xpu:0")
             else:
                 tokens = shared.tokenizer.encode(prompt, return_tensors='pt').cuda()
             output = shared.model(input_ids=tokens)

modules/models.py

@@ -7,7 +7,12 @@ from pathlib import Path
 
 import torch
 import transformers
-from accelerate import infer_auto_device_map, init_empty_weights
+from accelerate import (
+    infer_auto_device_map,
+    init_empty_weights,
+    is_ccl_available,
+    is_xpu_available
+)
 from transformers import (
     AutoConfig,
     AutoModel,
@@ -38,6 +43,10 @@ if shared.args.deepspeed:
     # Distributed setup
     local_rank = shared.args.local_rank if shared.args.local_rank is not None else int(os.getenv("LOCAL_RANK", "0"))
     world_size = int(os.getenv("WORLD_SIZE", "1"))
+    if is_xpu_available() and is_ccl_available():
+        torch.xpu.set_device(local_rank)
+        deepspeed.init_distributed(backend="ccl")
+    else:
         torch.cuda.set_device(local_rank)
         deepspeed.init_distributed()
     ds_config = generate_ds_config(shared.args.bf16, 1 * world_size, shared.args.nvme_offload_dir)
@@ -137,8 +146,9 @@ def huggingface_loader(model_name):
         if torch.backends.mps.is_available():
             device = torch.device('mps')
             model = model.to(device)
-        elif hasattr(torch, 'xpu') and torch.xpu.is_available():
-            model = model.to('xpu')
+        elif is_xpu_available():
+            device = torch.device("xpu")
+            model = model.to(device)
         else:
             model = model.cuda()
 
@@ -151,15 +161,10 @@ def huggingface_loader(model_name):
 
     # Load with quantization and/or offloading
     else:
-        conditions = [
-            shared.args.cpu,
-            torch.cuda.is_available(),
-            torch.backends.mps.is_available(),
-            hasattr(torch, 'xpu') and torch.xpu.is_available(),
-        ]
 
-        if not any(conditions):
-            logger.warning('No GPU has been detected by Pytorch. Falling back to CPU mode.')
+        if not any((shared.args.cpu, torch.cuda.is_available(), is_xpu_available(), torch.backends.mps.is_available())):
+            logger.warning('torch.cuda.is_available() and is_xpu_available() returned False. This means that no GPU has been detected. Falling back to CPU mode.')
+
             shared.args.cpu = True
 
         if shared.args.cpu:
@@ -362,7 +367,12 @@ def RWKV_loader(model_name):
     '''
     from modules.RWKV import RWKVModel, RWKVTokenizer
 
-    model = RWKVModel.from_pretrained(Path(f'{shared.args.model_dir}/{model_name}'), dtype="fp32" if shared.args.cpu else "bf16" if shared.args.bf16 else "fp16", device="cpu" if shared.args.cpu else "cuda")
+    model = RWKVModel.from_pretrained(
+        Path(f'{shared.args.model_dir}/{model_name}'),
+        dtype="fp32" if shared.args.cpu else "bf16" if shared.args.bf16 else "fp16",
+        device="cpu" if shared.args.cpu else "xpu" if is_xpu_available() else "cuda"
+    )
+
     tokenizer = RWKVTokenizer.from_pretrained(Path(shared.args.model_dir))
     return model, tokenizer
 
@@ -380,6 +390,9 @@ def get_max_memory_dict():
     # If --auto-devices is provided standalone, try to get a reasonable value
     # for the maximum memory of device :0
     elif shared.args.auto_devices:
+        if is_xpu_available():
+            total_mem = (torch.xpu.get_device_properties(0).total_memory / (1024 * 1024))
+        else:
             total_mem = (torch.cuda.get_device_properties(0).total_memory / (1024 * 1024))
         suggestion = round((total_mem - 1000) / 1000) * 1000
         if total_mem - suggestion < 800:
@@ -395,6 +408,9 @@ def get_max_memory_dict():
 def clear_torch_cache():
     gc.collect()
     if not shared.args.cpu:
+        if is_xpu_available():
+            torch.xpu.empty_cache()
+        else:
             torch.cuda.empty_cache()
 
 

modules/sampler_hijack.py

@@ -2,7 +2,7 @@ import math
 
 import torch
 import transformers
-from transformers import LogitsWarper
+from transformers import LogitsWarper, is_torch_xpu_available
 from transformers.generation.logits_process import (
     LogitNormalization,
     LogitsProcessor,
@@ -106,8 +106,11 @@ class MirostatLogitsWarper(LogitsWarper):
                 break
 
         # Normalize the probabilities of the remaining words
+        if is_torch_xpu_available():
+            prob_topk = torch.softmax(sorted_logits, dim=0).to("xpu")
+            prev_i = torch.multinomial(prob_topk, num_samples=1, replacement=True).to("xpu")
+        else:
             prob_topk = torch.softmax(sorted_logits, dim=0).to('cuda')
-
             prev_i = torch.multinomial(prob_topk, num_samples=1, replacement=True).to('cuda')
 
         observed_surprise = -math.log2(prob_topk[prev_i])

modules/text_generation.py

@@ -9,7 +9,7 @@ import traceback
 import numpy as np
 import torch
 import transformers
-from transformers import LogitsProcessorList
+from transformers import LogitsProcessorList, is_torch_xpu_available
 
 import modules.shared as shared
 from modules.callbacks import (
@@ -132,8 +132,8 @@ def encode(prompt, add_special_tokens=True, add_bos_token=True, truncation_lengt
     elif torch.backends.mps.is_available():
         device = torch.device('mps')
         return input_ids.to(device)
-    elif hasattr(torch, 'xpu') and torch.xpu.is_available():
-        return input_ids.to('xpu')
+    elif is_torch_xpu_available():
+        return input_ids.to("xpu:0")
     else:
         return input_ids.cuda()
 
@@ -238,7 +238,8 @@ def set_manual_seed(seed):
     torch.manual_seed(seed)
     if torch.cuda.is_available():
         torch.cuda.manual_seed_all(seed)
-
+    elif is_torch_xpu_available():
+        torch.xpu.manual_seed_all(seed)
     return seed
 
 

modules/training.py

@@ -26,6 +26,7 @@ from peft import (
 )
 from peft.utils.other import \
     TRANSFORMERS_MODELS_TO_LORA_TARGET_MODULES_MAPPING as model_to_lora_modules
+from transformers import is_torch_xpu_available
 from transformers.models.auto.modeling_auto import (
     MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
 )
@@ -626,6 +627,7 @@ def do_train(lora_name: str, always_override: bool, q_proj_en: bool, v_proj_en:
             # TODO: Enable multi-device support
             ddp_find_unused_parameters=None,
             no_cuda=shared.args.cpu,
+            use_ipex=True if is_torch_xpu_available and not shared.args.cpu else False
         ),
         data_collator=transformers.DataCollatorForLanguageModeling(shared.tokenizer, mlm=False),
         callbacks=list([Callbacks()])

modules/ui.py

@@ -4,10 +4,10 @@ from pathlib import Path
 import gradio as gr
 import torch
 import yaml
+from transformers import is_torch_xpu_available
 
 from modules import shared
 
-
 with open(Path(__file__).resolve().parent / '../css/NotoSans/stylesheet.css', 'r') as f:
     css = f.read()
 with open(Path(__file__).resolve().parent / '../css/main.css', 'r') as f:
@@ -85,7 +85,10 @@ def list_model_elements():
         'rope_freq_base',
         'numa',
     ]
-
+    if is_torch_xpu_available():
+        for i in range(torch.xpu.device_count()):
+            elements.append(f'gpu_memory_{i}')
+    else:
         for i in range(torch.cuda.device_count()):
             elements.append(f'gpu_memory_{i}')
 

modules/ui_model_menu.py

@@ -8,6 +8,7 @@ from pathlib import Path
 import gradio as gr
 import psutil
 import torch
+from transformers import is_torch_xpu_available
 
 from modules import loaders, shared, ui, utils
 from modules.logging_colors import logger
@@ -27,6 +28,10 @@ def create_ui():
 
     # Finding the default values for the GPU and CPU memories
     total_mem = []
+    if is_torch_xpu_available():
+        for i in range(torch.xpu.device_count()):
+            total_mem.append(math.floor(torch.xpu.get_device_properties(i).total_memory / (1024 * 1024)))
+    else:
         for i in range(torch.cuda.device_count()):
             total_mem.append(math.floor(torch.cuda.get_device_properties(i).total_memory / (1024 * 1024)))
 

one_click.py

@@ -56,6 +56,19 @@ def cpu_has_avx2():
         return True
 
 
+def cpu_has_amx():
+    try:
+        import cpuinfo
+
+        info = cpuinfo.get_cpu_info()
+        if 'amx' in info['flags']:
+            return True
+        else:
+            return False
+    except:
+        return True
+
+
 def torch_version():
     site_packages_path = None
     for sitedir in site.getsitepackages():