Upload folder using huggingface_hub

quanto-exps-2/benchmark.py

@@ -0,0 +1,135 @@
+import argparse
+from optimum.quanto import freeze, qfloat8, qint4, qint8, quantize
+import torch
+import json
+import torch.utils.benchmark as benchmark
+from diffusers import DiffusionPipeline
+import gc
+
+
+WARM_UP_ITERS = 5
+PROMPT = "ghibli style, a fantasy landscape with castles"
+
+TORCH_DTYPES = {"fp32": torch.float32, "fp16": torch.float16, "bf16": torch.bfloat16}
+# QTYPES = {"fp8": qfloat8, "int8": qint8, "int4": qint4, "none": None}
+QTYPES = {"fp8": qfloat8, "int8": qint8, "none": None}
+
+PREFIXES = {
+    "stabilityai/stable-diffusion-3-medium-diffusers": "sd3",
+    "PixArt-alpha/PixArt-Sigma-XL-2-1024-MS": "pixart",
+    "fal/AuraFlow": "auraflow",
+}
+
+def flush():
+    """Wipes off memory."""
+    gc.collect()
+    torch.cuda.empty_cache()
+    torch.cuda.reset_max_memory_allocated()
+    torch.cuda.reset_peak_memory_stats()
+
+def load_pipeline(ckpt_id, torch_dtype, qtype=None, exclude_layers=None, qte=False, fuse=False):
+    pipe = DiffusionPipeline.from_pretrained(ckpt_id, torch_dtype=torch_dtype).to("cuda")
+    if fuse:
+        pipe.transformer.fuse_qkv_projections()
+
+    if qtype:
+        quantize(pipe.transformer, weights=qtype, exclude=exclude_layers)
+        freeze(pipe.transformer)
+
+        if qte:
+            quantize(pipe.text_encoder, weights=qtype)
+            freeze(pipe.text_encoder)
+            if hasattr(pipe, "text_encoder_2"):
+                quantize(pipe.text_encoder_2, weights=qtype)
+                freeze(pipe.text_encoder_2)
+            if hasattr(pipe, "text_encoder_3"):
+                quantize(pipe.text_encoder_3, weights=qtype)
+                freeze(pipe.text_encoder_3)
+
+    pipe.set_progress_bar_config(disable=True)
+    return pipe
+
+
+def run_inference(pipe, batch_size=1):
+    _ = pipe(
+        prompt=PROMPT,
+        num_images_per_prompt=batch_size,
+        generator=torch.manual_seed(0),
+    )
+
+
+def benchmark_fn(f, *args, **kwargs):
+    t0 = benchmark.Timer(stmt="f(*args, **kwargs)", globals={"args": args, "kwargs": kwargs, "f": f})
+    return f"{(t0.blocked_autorange().mean):.3f}"
+
+
+def bytes_to_giga_bytes(bytes):
+    return f"{(bytes / 1024 / 1024 / 1024):.3f}"
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--ckpt_id",
+        type=str,
+        default="stabilityai/stable-diffusion-3-medium-diffusers",
+        choices=list(PREFIXES.keys()),
+    )
+    parser.add_argument("--batch_size", type=int, default=1)
+    parser.add_argument("--torch_dtype", type=str, default="fp16", choices=list(TORCH_DTYPES.keys()))
+    parser.add_argument("--qtype", type=str, default="none", choices=list(QTYPES.keys()))
+    parser.add_argument("--qte", type=int, default=0, help="Quantize text encoder")
+    parser.add_argument("--fuse", type=int, default=0)
+    parser.add_argument("--exclude_layers", metavar="N", type=str, nargs="*", default=None)
+    args = parser.parse_args()
+
+    flush()
+
+    print(
+        f"Running with ckpt_id: {args.ckpt_id}, batch_size: {args.batch_size}, torch_dtype: {args.torch_dtype}, qtype: {args.qtype}, qte: {bool(args.qte)}, {args.exclude_layers=}, {args.fuse=}"
+    )
+    pipeline = load_pipeline(
+        ckpt_id=args.ckpt_id,
+        torch_dtype=TORCH_DTYPES[args.torch_dtype],
+        qtype=QTYPES[args.qtype],
+        exclude_layers=args.exclude_layers,
+        qte=args.qte,
+        fuse=bool(args.fuse),
+    )
+
+    for _ in range(WARM_UP_ITERS):
+        run_inference(pipeline, args.batch_size)
+
+    time = benchmark_fn(run_inference, pipeline, args.batch_size)
+    torch.cuda.empty_cache()
+    memory = bytes_to_giga_bytes(torch.cuda.memory_allocated())  # in GBs.
+    print(
+        f"ckpt: {args.ckpt_id} batch_size: {args.batch_size}, qte: {args.qte}, {args.exclude_layers=} "
+        f"torch_dtype: {args.torch_dtype}, qtype: {args.qtype}  in {time} seconds and {memory} GBs."
+    )
+
+    ckpt_id = PREFIXES[args.ckpt_id]
+    img_name = f"ckpt@{ckpt_id}-bs@{args.batch_size}-dtype@{args.torch_dtype}-qtype@{args.qtype}-qte@{args.qte}-fuse@{args.fuse}"
+    if args.exclude_layers:
+        exclude_layers = "_".join(args.exclude_layers)
+        img_name += f"-exclude@{exclude_layers}"
+    image = pipeline(
+        prompt=PROMPT,
+        num_images_per_prompt=args.batch_size,
+        generator=torch.manual_seed(0),
+    ).images[0]
+    image.save(f"{img_name}.png")
+
+    info = dict(
+        batch_size=args.batch_size,
+        memory=memory,
+        time=time,
+        dtype=args.torch_dtype,
+        qtype=args.qtype,
+        qte=args.qte,
+        exclude_layers=args.exclude_layers,
+        # fuse=args.fuse,
+    )
+    info_file = f"{img_name}_info.json"
+    with open(info_file, "w") as f:
+        json.dump(info, f)

quanto-exps-2/benchmark_2.py

@@ -0,0 +1,147 @@
+import argparse
+from optimum.quanto import freeze, qfloat8, qint4, qint8, quantize
+import torch
+import json
+import torch.utils.benchmark as benchmark
+from diffusers import DiffusionPipeline
+import gc
+
+
+WARM_UP_ITERS = 5
+PROMPT = "ghibli style, a fantasy landscape with castles"
+
+TORCH_DTYPES = {"fp32": torch.float32, "fp16": torch.float16, "bf16": torch.bfloat16}
+QTYPES = {"fp8": qfloat8, "int8": qint8, "int4": qint4, "none": None}
+
+PREFIXES = {
+    "stabilityai/stable-diffusion-3-medium-diffusers": "sd3",
+}
+
+
+def flush():
+    """Wipes off memory."""
+    gc.collect()
+    torch.cuda.empty_cache()
+    torch.cuda.reset_max_memory_allocated()
+    torch.cuda.reset_peak_memory_stats()
+
+
+def load_pipeline(
+    ckpt_id, torch_dtype, qtype=None, exclude_layers=None, qte=False, first=False, second=False, third=False
+):
+    pipe = DiffusionPipeline.from_pretrained(ckpt_id, torch_dtype=torch_dtype).to("cuda")
+
+    if qtype:
+        quantize(pipe.transformer, weights=qtype, exclude=exclude_layers)
+        freeze(pipe.transformer)
+
+        if qte:
+            if first:
+                quantize(pipe.text_encoder, weights=qtype)
+                freeze(pipe.text_encoder)
+            if second:
+                quantize(pipe.text_encoder_2, weights=qtype)
+                freeze(pipe.text_encoder)
+            if third:
+                quantize(pipe.text_encoder_3, weights=qtype)
+                freeze(pipe.text_encoder_3)
+
+    pipe.set_progress_bar_config(disable=True)
+    return pipe
+
+
+def run_inference(pipe, batch_size=1):
+    _ = pipe(
+        prompt=PROMPT,
+        num_images_per_prompt=batch_size,
+        generator=torch.manual_seed(0),
+    )
+
+
+def benchmark_fn(f, *args, **kwargs):
+    t0 = benchmark.Timer(stmt="f(*args, **kwargs)", globals={"args": args, "kwargs": kwargs, "f": f})
+    return f"{(t0.blocked_autorange().mean):.3f}"
+
+
+def bytes_to_giga_bytes(bytes):
+    return f"{(bytes / 1024 / 1024 / 1024):.3f}"
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--ckpt_id",
+        type=str,
+        default="stabilityai/stable-diffusion-3-medium-diffusers",
+        choices=list(PREFIXES.keys()),
+    )
+    parser.add_argument("--batch_size", type=int, default=1)
+    parser.add_argument("--torch_dtype", type=str, default="fp16", choices=list(TORCH_DTYPES.keys()))
+    parser.add_argument("--qtype", type=str, default="none", choices=list(QTYPES.keys()))
+    parser.add_argument("--qte", type=int, default=0, help="Quantize text encoder")
+    parser.add_argument("--first", type=int, default=0, help="Quantize first text encoder")
+    parser.add_argument("--second", type=int, default=0, help="Quantize second text encoder")
+    parser.add_argument("--third", type=int, default=0, help="Quantize third text encoder")
+    parser.add_argument("--exclude_layers", metavar="N", type=str, nargs="*", default=None)
+    args = parser.parse_args()
+
+    flush()
+
+    print(
+        f"Running with ckpt_id: {args.ckpt_id}, batch_size: {args.batch_size}, torch_dtype: {args.torch_dtype}, qtype: {args.qtype}, qte: {bool(args.qte)}"
+    )
+    pipeline = load_pipeline(
+        ckpt_id=args.ckpt_id,
+        torch_dtype=TORCH_DTYPES[args.torch_dtype],
+        qtype=QTYPES[args.qtype],
+        exclude_layers=args.exclude_layers,
+        qte=args.qte,
+        first=args.first,
+        second=args.second,
+        third=args.third,
+    )
+
+    for _ in range(WARM_UP_ITERS):
+        run_inference(pipeline, args.batch_size)
+
+    time = benchmark_fn(run_inference, pipeline, args.batch_size)
+    memory = bytes_to_giga_bytes(torch.cuda.max_memory_allocated())  # in GBs.
+    print(
+        f"ckpt: {args.ckpt_id} batch_size: {args.batch_size}, qte: {args.qte}, "
+        f"torch_dtype: {args.torch_dtype}, qtype: {args.qtype}  in {time} seconds and {memory} GBs."
+    )
+
+    ckpt_id = PREFIXES[args.ckpt_id]
+    img_name = f"ckpt@{ckpt_id}-bs@{args.batch_size}-dtype@{args.torch_dtype}-qtype@{args.qtype}-qte@{args.qte}"
+    if args.exclude_layers:
+        exclude_layers = "_".join(args.exclude_layers)
+        img_name += f"-exclude@{exclude_layers}"
+    if args.first:
+        img_name += f"-first@{args.first}"
+    if args.second:
+        img_name += f"-second@{args.second}"
+    if args.third:
+        img_name += f"-third@{args.third}"
+
+    image = pipeline(
+        prompt=PROMPT,
+        num_images_per_prompt=args.batch_size,
+        generator=torch.manual_seed(0),
+    ).images[0]
+    image.save(f"{img_name}.png")
+
+    info = dict(
+        batch_size=args.batch_size,
+        memory=memory,
+        time=time,
+        dtype=args.torch_dtype,
+        qtype=args.qtype,
+        qte=args.qte,
+        exclude_layers=args.exclude_layers,
+        first=args.first,
+        second=args.second,
+        third=args.third,
+    )
+    info_file = f"{img_name}_info.json"
+    with open(info_file, "w") as f:
+        json.dump(info, f)

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@fp8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "8.861", "time": "14.298", "dtype": "bf16", "qtype": "fp8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@fp8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "7.783", "time": "14.345", "dtype": "bf16", "qtype": "fp8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@int8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "8.860", "time": "13.769", "dtype": "bf16", "qtype": "int8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@int8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "7.783", "time": "13.779", "dtype": "bf16", "qtype": "int8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@none-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "15.219", "time": "12.468", "dtype": "bf16", "qtype": "none", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@bf16-qtype@none-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "15.219", "time": "12.485", "dtype": "bf16", "qtype": "none", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@fp8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "9.466", "time": "14.629", "dtype": "fp16", "qtype": "fp8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@fp8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "7.920", "time": "14.624", "dtype": "fp16", "qtype": "fp8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@int8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "9.466", "time": "14.021", "dtype": "fp16", "qtype": "int8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@int8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "7.921", "time": "14.031", "dtype": "fp16", "qtype": "int8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@none-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "15.821", "time": "12.666", "dtype": "fp16", "qtype": "none", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@1-dtype@fp16-qtype@none-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 1, "memory": "15.821", "time": "12.648", "dtype": "fp16", "qtype": "none", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@fp8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "8.860", "time": "50.247", "dtype": "bf16", "qtype": "fp8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@fp8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "7.783", "time": "50.217", "dtype": "bf16", "qtype": "fp8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@int8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "8.860", "time": "49.705", "dtype": "bf16", "qtype": "int8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@int8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "7.784", "time": "49.671", "dtype": "bf16", "qtype": "int8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@none-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "15.218", "time": "48.483", "dtype": "bf16", "qtype": "none", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@bf16-qtype@none-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "15.219", "time": "48.437", "dtype": "bf16", "qtype": "none", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@fp8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "9.467", "time": "51.857", "dtype": "fp16", "qtype": "fp8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@fp8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "7.921", "time": "51.903", "dtype": "fp16", "qtype": "fp8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@int8-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "9.467", "time": "51.386", "dtype": "fp16", "qtype": "int8", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@int8-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "7.922", "time": "51.283", "dtype": "fp16", "qtype": "int8", "qte": 1, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@none-qte@0-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "15.819", "time": "49.271", "dtype": "fp16", "qtype": "none", "qte": 0, "exclude_layers": null}

quanto-exps-2/ckpt@auraflow-bs@4-dtype@fp16-qtype@none-qte@1-fuse@0_info.json

@@ -0,0 +1 @@
+{"batch_size": 4, "memory": "15.819", "time": "49.254", "dtype": "fp16", "qtype": "none", "qte": 1, "exclude_layers": null}