[LCM] Better error message

+

@@ -0,0 +1,21 @@
+#!/usr/bin/env python3
+from diffusers import AutoPipelineForText2Image, AutoPipelineForImage2Image
+import torch
+
+pipe = AutoPipelineForText2Image.from_pretrained("SimianLuo/LCM_Dreamshaper_v7", dtype=torch.float16)
+
+prompt = "A beautiful landscape of a snowy mountain"
+
+num_inference_steps = 1
+
+image = pipe(prompt=prompt, num_inference_steps=num_inference_steps, output_type="pil").images[0]
+image.save("snow_mountain.png")
+
+pipe = AutoPipelineForImage2Image.from_pipe(pipe)
+
+prompt = "A beautiful landscape of a very red mountain"
+
+image = pipe(prompt=prompt, image=image, num_inference_steps=10, strength=0.05, output_type="pil").images[0]
+image.show()
+
+

bug_5776.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+from diffusers import AutoPipelineForText2Image
+import torch
+
+pipeline = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16).to("cuda")
+pipeline.load_lora_weights("ostris/super-cereal-sdxl-lora", weight_name="cereal_box_sdxl_v1.safetensors")
+print(pipeline.unet.attn_processor)

clean.py

@@ -0,0 +1,2 @@
+#!/usr/bin/env python3
+from diffusers import *

delete_function.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python3
+import sys
+
+filenames = sys.argv[1:]
+
+MATCH_PATTERN_1 = "# Copied from transformers.models.bart.modeling_bart._make_causal_mask"
+MATCH_PATTERN_2 = "def _make_causal_mask("
+
+MATCH_PATTERN_1 = "# Copied from transformers.models.bart.modeling_bart.prepare_4d_attention_mask"
+MATCH_PATTERN_2 = "def prepare_4d_attention_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None):"
+
+END_MATCH_PATTERN_2 = ""
+
+# MATCH_PATTERN_1 = "def _prepare_decoder_attention_mask(self, attention_mask, input_shape, inputs_embeds, past_key_values_length):"
+#MATCH_PATTERN_2 = "# create causal mask"
+
+# END_MATCH_PATTERN_2 = "def forward("
+
+                                        
+for filename in filenames:
+    with open(filename, "r") as f:
+        lines = f.readlines()
+
+    new_lines = []
+    is_in_del = False
+    for i, line in enumerate(lines):
+        if line.strip().lstrip() == MATCH_PATTERN_1 and i < len(lines) - 1 and lines[i + 1].strip().lstrip() == MATCH_PATTERN_2:
+            print("suh")
+            is_in_del = True
+        elif line.strip().lstrip() == "" and i < len(lines) - 1 and lines[i + 1].strip().lstrip() == END_MATCH_PATTERN_2:
+            is_in_del = False
+
+        if not is_in_del:
+            new_lines.append(line)
+
+
+    with open(filename, "w") as f:
+        f.writelines(new_lines)

run_lcm.py

@@ -0,0 +1,21 @@
+#!/usr/bin/env python3
+from diffusers import AutoPipelineForText2Image, AutoPipelineForImage2Image
+import torch
+
+pipe = AutoPipelineForText2Image.from_pretrained("SimianLuo/LCM_Dreamshaper_v7", dtype=torch.float16)
+
+prompt = "A beautiful landscape of a snowy mountain"
+
+num_inference_steps = 1
+
+image = pipe(prompt=prompt, num_inference_steps=num_inference_steps, output_type="pil").images[0]
+image.save("snow_mountain.png")
+
+pipe = AutoPipelineForImage2Image.from_pipe(pipe)
+
+prompt = "A beautiful landscape of a very red mountain"
+
+image = pipe(prompt=prompt, image=image, num_inference_steps=10, strength=0.05, output_type="pil").images[0]
+image.show()
+
+

run_randn.py

@@ -0,0 +1,13 @@
+#!/usr/bin/env python3
+import torch
+
+def get_random():
+    return torch.randn((2,2))
+
+seeds = []
+for _ in range(5):
+    num = get_random()
+    seeds.append(torch.initial_seed())
+    print(num)
+
+print("Seeds should be different, but are not", seeds)

run_whisper.py

@@ -0,0 +1,20 @@
+#!/usr/bin/env python3
+from transformers import WhisperForCausalLM, WhisperForConditionalGeneration, WhisperProcessor
+import torch
+from datasets import load_dataset
+
+processor = WhisperProcessor.from_pretrained("openai/whisper-large-v2")
+model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-large-v2")
+
+assistant_model = WhisperForCausalLM.from_pretrained("patrickvonplaten/whisper-large-v2-32-2")
+
+ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
+sample = ds[0]["audio"]
+input_features = processor(sample["array"], sampling_rate=sample["sampling_rate"], return_tensors="pt").input_features 
+
+predicted_ids = model.generate(input_features, assistant_model=assistant_model)
+
+# decode token ids to text
+transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
+
+print(transcription)