model: add glm-asr support #17901
Open
model: add glm-asr support #17901
+156
−19
Conversation
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
ngxson
requested changes
Dec 10, 2025
CISC
reviewed
Dec 10, 2025
convert_hf_to_gguf.py
Outdated
Comment on lines
2414
to
2423
| if isinstance(self.hparams.get("eos_token_id"), list): | ||
| from transformers import AutoTokenizer | ||
| tokenizer = AutoTokenizer.from_pretrained(self.dir_model, trust_remote_code=True) | ||
| special_vocab = gguf.SpecialVocab(self.dir_model, load_merges=True) | ||
| special_vocab._set_special_token("eos", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab._set_special_token("eot", tokenizer.get_added_vocab()["<|user|>"]) | ||
| special_vocab._set_special_token("unk", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab._set_special_token("bos", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab.add_to_gguf(self.gguf_writer) | ||
| special_vocab.chat_template = "glmedge" |
Collaborator
There was a problem hiding this comment.
This is not ok, check for root architecture instead (see Qwen3MoeModel), also you should not need to set any of those special tokens.
Also, setting the template name like that doesn't work any more I think, and it's a dirty hack to begin with, if the model creators can't be bothered, neither should we.
ngxson
reviewed
Dec 10, 2025
…build_stack for padding and review
ngxson
reviewed
Dec 12, 2025
Comment on lines
+2569
to
+2578
| def set_vocab(self): | ||
| super().set_vocab() | ||
| from transformers import AutoTokenizer | ||
| tokenizer = AutoTokenizer.from_pretrained(self.dir_model, trust_remote_code=True) | ||
| special_vocab = gguf.SpecialVocab(self.dir_model, load_merges=True) | ||
| special_vocab._set_special_token("eos", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab._set_special_token("eot", tokenizer.get_added_vocab()["<|user|>"]) | ||
| special_vocab._set_special_token("unk", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab._set_special_token("bos", tokenizer.get_added_vocab()["<|endoftext|>"]) | ||
| special_vocab.add_to_gguf(self.gguf_writer) |
Collaborator
There was a problem hiding this comment.
It seems like the original comment was not addressed: #17901 (comment)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Make sure to read the contributing guidelines before submitting a PR
This PR adds support for the GLM-ASR architecture, specifically validating with the zai-org/GLM-ASR-Nano-2512 model.
Key Changes:
convert_hf_to_gguf.pyto handle dynamic configuration keys (glm-asr use "lm_config" instead of text_config). It now correctly identifies the config section by checking:llm_config_key = "lm_config" if "lm_config" in self.hparams else "text_config"Result