Commit ·
65631dc
1
Parent(s): d42ff0c
docs: update volume mount syntax to hf:// prefix
Browse fileshf jobs uv run -v now requires hf://[TYPE/]SOURCE:/MOUNT_PATH.
The old bare `bucket/user/name` form errors with
"Invalid volume format: Source must start with 'hf://'".
Updates README Quick Start and all four script docstrings /
print helpers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- README.md +4 -4
- cohere-transcribe-vllm.py +6 -6
- cohere-transcribe.py +6 -6
- download-ia.py +3 -3
- easytranscriber-transcribe.py +6 -6
README.md
CHANGED
|
@@ -19,15 +19,15 @@ Scripts run directly from their Hub URL — no clone or local checkout needed:
|
|
| 19 |
```bash
|
| 20 |
# 1. Download audio from Internet Archive straight into a bucket
|
| 21 |
hf jobs uv run \
|
| 22 |
-
-v
|
| 23 |
https://huggingface.co/datasets/uv-scripts/transcription/raw/main/download-ia.py \
|
| 24 |
SUSPENSE /output
|
| 25 |
|
| 26 |
# 2. Transcribe — audio bucket in, transcript bucket out
|
| 27 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \
|
| 28 |
-
-e UV_TORCH_BACKEND=
|
| 29 |
-
-v
|
| 30 |
-
-v
|
| 31 |
https://huggingface.co/datasets/uv-scripts/transcription/raw/main/cohere-transcribe.py \
|
| 32 |
/input /output --language en --compile
|
| 33 |
```
|
|
|
|
| 19 |
```bash
|
| 20 |
# 1. Download audio from Internet Archive straight into a bucket
|
| 21 |
hf jobs uv run \
|
| 22 |
+
-v hf://buckets/user/audio-files:/output \
|
| 23 |
https://huggingface.co/datasets/uv-scripts/transcription/raw/main/download-ia.py \
|
| 24 |
SUSPENSE /output
|
| 25 |
|
| 26 |
# 2. Transcribe — audio bucket in, transcript bucket out
|
| 27 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \
|
| 28 |
+
-e UV_TORCH_BACKEND=cu128 \
|
| 29 |
+
-v hf://buckets/user/audio-files:/input:ro \
|
| 30 |
+
-v hf://buckets/user/transcripts:/output \
|
| 31 |
https://huggingface.co/datasets/uv-scripts/transcription/raw/main/cohere-transcribe.py \
|
| 32 |
/input /output --language en --compile
|
| 33 |
```
|
cohere-transcribe-vllm.py
CHANGED
|
@@ -37,8 +37,8 @@ Examples:
|
|
| 37 |
# HF Jobs with bucket volumes
|
| 38 |
hf jobs uv run --flavor l4x1 \\
|
| 39 |
-s HF_TOKEN \\
|
| 40 |
-
-v
|
| 41 |
-
-v
|
| 42 |
transcribe.py /input /output --language en
|
| 43 |
|
| 44 |
Model: CohereLabs/cohere-transcribe-03-2026 (2B, Apache 2.0)
|
|
@@ -141,8 +141,8 @@ Examples:
|
|
| 141 |
|
| 142 |
HF Jobs with bucket volumes:
|
| 143 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 144 |
-
-v
|
| 145 |
-
-v
|
| 146 |
transcribe.py /input /output --language en
|
| 147 |
""",
|
| 148 |
)
|
|
@@ -328,8 +328,8 @@ if __name__ == "__main__":
|
|
| 328 |
print()
|
| 329 |
print("HF Jobs with bucket volumes:")
|
| 330 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN \\")
|
| 331 |
-
print(" -v
|
| 332 |
-
print(" -v
|
| 333 |
print(" transcribe.py /input /output --language en")
|
| 334 |
print()
|
| 335 |
print("For full help: uv run transcribe.py --help")
|
|
|
|
| 37 |
# HF Jobs with bucket volumes
|
| 38 |
hf jobs uv run --flavor l4x1 \\
|
| 39 |
-s HF_TOKEN \\
|
| 40 |
+
-v hf://buckets/user/audio-input:/input:ro \\
|
| 41 |
+
-v hf://buckets/user/transcripts:/output \\
|
| 42 |
transcribe.py /input /output --language en
|
| 43 |
|
| 44 |
Model: CohereLabs/cohere-transcribe-03-2026 (2B, Apache 2.0)
|
|
|
|
| 141 |
|
| 142 |
HF Jobs with bucket volumes:
|
| 143 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 144 |
+
-v hf://buckets/user/audio-bucket:/input:ro \\
|
| 145 |
+
-v hf://buckets/user/transcripts:/output \\
|
| 146 |
transcribe.py /input /output --language en
|
| 147 |
""",
|
| 148 |
)
|
|
|
|
| 328 |
print()
|
| 329 |
print("HF Jobs with bucket volumes:")
|
| 330 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN \\")
|
| 331 |
+
print(" -v hf://buckets/user/audio-input:/input:ro \\")
|
| 332 |
+
print(" -v hf://buckets/user/transcripts:/output \\")
|
| 333 |
print(" transcribe.py /input /output --language en")
|
| 334 |
print()
|
| 335 |
print("For full help: uv run transcribe.py --help")
|
cohere-transcribe.py
CHANGED
|
@@ -31,8 +31,8 @@ Examples:
|
|
| 31 |
# HF Jobs with bucket volumes
|
| 32 |
hf jobs uv run --flavor l4x1 \\
|
| 33 |
-s HF_TOKEN \\
|
| 34 |
-
-v
|
| 35 |
-
-v
|
| 36 |
transcribe-transformers.py /input /output --language en --compile
|
| 37 |
|
| 38 |
Model: CohereLabs/cohere-transcribe-03-2026 (2B, Apache 2.0)
|
|
@@ -102,8 +102,8 @@ Examples:
|
|
| 102 |
|
| 103 |
HF Jobs with bucket volumes:
|
| 104 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 105 |
-
-v
|
| 106 |
-
-v
|
| 107 |
transcribe-transformers.py /input /output --language en --compile
|
| 108 |
""",
|
| 109 |
)
|
|
@@ -275,8 +275,8 @@ if __name__ == "__main__":
|
|
| 275 |
print()
|
| 276 |
print("HF Jobs with bucket volumes:")
|
| 277 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN \\")
|
| 278 |
-
print(" -v
|
| 279 |
-
print(" -v
|
| 280 |
print(" transcribe-transformers.py /input /output --language en --compile")
|
| 281 |
print()
|
| 282 |
print("For full help: uv run transcribe-transformers.py --help")
|
|
|
|
| 31 |
# HF Jobs with bucket volumes
|
| 32 |
hf jobs uv run --flavor l4x1 \\
|
| 33 |
-s HF_TOKEN \\
|
| 34 |
+
-v hf://buckets/user/audio-input:/input:ro \\
|
| 35 |
+
-v hf://buckets/user/transcripts:/output \\
|
| 36 |
transcribe-transformers.py /input /output --language en --compile
|
| 37 |
|
| 38 |
Model: CohereLabs/cohere-transcribe-03-2026 (2B, Apache 2.0)
|
|
|
|
| 102 |
|
| 103 |
HF Jobs with bucket volumes:
|
| 104 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 105 |
+
-v hf://buckets/user/audio-bucket:/input:ro \\
|
| 106 |
+
-v hf://buckets/user/transcripts:/output \\
|
| 107 |
transcribe-transformers.py /input /output --language en --compile
|
| 108 |
""",
|
| 109 |
)
|
|
|
|
| 275 |
print()
|
| 276 |
print("HF Jobs with bucket volumes:")
|
| 277 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN \\")
|
| 278 |
+
print(" -v hf://buckets/user/audio-input:/input:ro \\")
|
| 279 |
+
print(" -v hf://buckets/user/transcripts:/output \\")
|
| 280 |
print(" transcribe-transformers.py /input /output --language en --compile")
|
| 281 |
print()
|
| 282 |
print("For full help: uv run transcribe-transformers.py --help")
|
download-ia.py
CHANGED
|
@@ -18,7 +18,7 @@ Examples:
|
|
| 18 |
|
| 19 |
# HF Jobs — download straight into a bucket
|
| 20 |
hf jobs uv run \
|
| 21 |
-
-v
|
| 22 |
download-ia.py SUSPENSE /output
|
| 23 |
|
| 24 |
# With filters
|
|
@@ -51,7 +51,7 @@ Examples:
|
|
| 51 |
|
| 52 |
HF Jobs with bucket volume:
|
| 53 |
hf jobs uv run \\
|
| 54 |
-
-v
|
| 55 |
download-ia.py SUSPENSE /output
|
| 56 |
""",
|
| 57 |
)
|
|
@@ -143,7 +143,7 @@ if __name__ == "__main__":
|
|
| 143 |
print()
|
| 144 |
print("HF Jobs with bucket:")
|
| 145 |
print(" hf jobs uv run \\")
|
| 146 |
-
print(" -v
|
| 147 |
print(" download-ia.py SUSPENSE /output")
|
| 148 |
print()
|
| 149 |
print("For full help: uv run download-ia.py --help")
|
|
|
|
| 18 |
|
| 19 |
# HF Jobs — download straight into a bucket
|
| 20 |
hf jobs uv run \
|
| 21 |
+
-v hf://buckets/user/audio-bucket:/output \
|
| 22 |
download-ia.py SUSPENSE /output
|
| 23 |
|
| 24 |
# With filters
|
|
|
|
| 51 |
|
| 52 |
HF Jobs with bucket volume:
|
| 53 |
hf jobs uv run \\
|
| 54 |
+
-v hf://buckets/user/audio-bucket:/output \\
|
| 55 |
download-ia.py SUSPENSE /output
|
| 56 |
""",
|
| 57 |
)
|
|
|
|
| 143 |
print()
|
| 144 |
print("HF Jobs with bucket:")
|
| 145 |
print(" hf jobs uv run \\")
|
| 146 |
+
print(" -v hf://buckets/user/audio-bucket:/output \\")
|
| 147 |
print(" download-ia.py SUSPENSE /output")
|
| 148 |
print()
|
| 149 |
print("For full help: uv run download-ia.py --help")
|
easytranscriber-transcribe.py
CHANGED
|
@@ -54,8 +54,8 @@ Examples:
|
|
| 54 |
# HF Jobs with bucket volumes
|
| 55 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 56 |
-e UV_TORCH_BACKEND=cu128 \\
|
| 57 |
-
-v
|
| 58 |
-
-v
|
| 59 |
easytranscriber-transcribe.py /input /output \\
|
| 60 |
--language en --emit-txt --emit-srt
|
| 61 |
"""
|
|
@@ -195,8 +195,8 @@ Examples:
|
|
| 195 |
|
| 196 |
HF Jobs with bucket volumes:
|
| 197 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN -e UV_TORCH_BACKEND=cu128 \\
|
| 198 |
-
-v
|
| 199 |
-
-v
|
| 200 |
easytranscriber-transcribe.py /input /output --language en --emit-txt --emit-srt
|
| 201 |
""",
|
| 202 |
)
|
|
@@ -418,8 +418,8 @@ if __name__ == "__main__":
|
|
| 418 |
print()
|
| 419 |
print("HF Jobs with bucket volumes:")
|
| 420 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN -e UV_TORCH_BACKEND=cu128 \\")
|
| 421 |
-
print(" -v
|
| 422 |
-
print(" -v
|
| 423 |
print(" easytranscriber-transcribe.py /input /output \\")
|
| 424 |
print(" --language en --emit-txt --emit-srt")
|
| 425 |
print()
|
|
|
|
| 54 |
# HF Jobs with bucket volumes
|
| 55 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN \\
|
| 56 |
-e UV_TORCH_BACKEND=cu128 \\
|
| 57 |
+
-v hf://buckets/user/audio-files:/input:ro \\
|
| 58 |
+
-v hf://buckets/user/transcripts-aligned:/output \\
|
| 59 |
easytranscriber-transcribe.py /input /output \\
|
| 60 |
--language en --emit-txt --emit-srt
|
| 61 |
"""
|
|
|
|
| 195 |
|
| 196 |
HF Jobs with bucket volumes:
|
| 197 |
hf jobs uv run --flavor l4x1 -s HF_TOKEN -e UV_TORCH_BACKEND=cu128 \\
|
| 198 |
+
-v hf://buckets/user/audio-files:/input:ro \\
|
| 199 |
+
-v hf://buckets/user/transcripts-aligned:/output \\
|
| 200 |
easytranscriber-transcribe.py /input /output --language en --emit-txt --emit-srt
|
| 201 |
""",
|
| 202 |
)
|
|
|
|
| 418 |
print()
|
| 419 |
print("HF Jobs with bucket volumes:")
|
| 420 |
print(" hf jobs uv run --flavor l4x1 -s HF_TOKEN -e UV_TORCH_BACKEND=cu128 \\")
|
| 421 |
+
print(" -v hf://buckets/user/audio-files:/input:ro \\")
|
| 422 |
+
print(" -v hf://buckets/user/transcripts-aligned:/output \\")
|
| 423 |
print(" easytranscriber-transcribe.py /input /output \\")
|
| 424 |
print(" --language en --emit-txt --emit-srt")
|
| 425 |
print()
|