Repository navigation
embeddings: header on generate-embeddings.py, and nine more declared models - #189
Merged
Merged
Conversation
The header sets l4x1, a 1h timeout and the HF_TOKEN secret, so the command needs no job flags. The docstring examples drop them. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
google/embeddinggemma-2 failed to load without it: its processor needs torchvision. With it, the model embeds 100 IMDb rows and uses its registered document prompt. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
A model reached the page only as a script's default value, so generate-embeddings.py declared one model although it takes any sentence-transformers model through --model. TESTED_MODELS lists the models that ran a recipe in a smoke Job with checked output. Each gets its own entry, with --model set in the command. Nine embedding models are added after a 100-row smoke Job each. Declared models go from 30 to 39. nomic-embed-text-v1.5 failed in its own remote code and is not listed. Records are in maintenance/verifications.jsonl. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
generate-embeddings.pytakes any sentence-transformers model through--model, but the Run on Jobs Space declared only its default. A model reached the page only as a script's default value.Changes
generate-embeddings.pygets a[tool.hf-jobs]header (l4x1, 1h,HF_TOKEN). The page now says "command ready" and the command needs no job flags.torchvisiondependency.google/embeddinggemma-2failed to load without it.TESTED_MODELSintools/build-catalog.py. Models other than a script's default that ran the recipe in a smoke Job with checked output. Each gets its own entry, with--modelset in the command.Checks
l4x1. Each output has 100 rows, unit-length vectors, 100 distinct vectors and no NaN. The card names the flavor and the prompt.passage:, embeddinggemma-2 gets its registeredtitle: none | text:, the others get none.maintenance/verifications.jsonl, andmaintenance/STATUS.mdis regenerated.python3 tools/build-catalog.py --verify-modelsandpython3 tools/recipe-status.py --checkpass.Not included
nomic-ai/nomic-embed-text-v1.5fails in its own remote code on the installedtransformers(get_extended_attention_mask). It is not listed. The recipe's prefix table still names nomic.🤖 Generated with Claude Code