初始化项目,由ModelHub XC社区提供模型

Model: jadael/comma-v0.1-2t-GGUF
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-12 04:14:16 +08:00
commit f42e3a0114
8 changed files with 658 additions and 0 deletions

23
comma_tokenizer.patch Normal file
View File

@@ -0,0 +1,23 @@
--- convert_hf_to_gguf.py.original 2025-10-18 05:00:00.000000000
+++ convert_hf_to_gguf.py.patched 2025-10-18 06:00:00.000000000
@@ -897,6 +897,10 @@
if chkhsh == "53e325976a6e142379c19b09afcae354f2f496f147afa8f9e189a33fe4e3024e":
# ref: https://huggingface.co/ibm-granite/granite-docling-258M
res = "granite-docling"
+ if chkhsh == "bf66900d65fe80247e435184a4ac839c5c332657cf567e64b8ede5fbd63f5fd9":
+ # Patch for Comma v0.1 tokenizer (Llama 3 compatible)
+ # ref: https://huggingface.co/common-pile/comma-v0.1-2t
+ res = "llama-bpe"
if res is None:
logger.warning("\n")
@@ -911,7 +915,7 @@
logger.warning(f"** chkhsh: {chkhsh}")
logger.warning("**************************************************************************************")
logger.warning("\n")
-
- raise NotImplementedError("BPE pre-tokenizer was not recognized - update get_vocab_base_pre()")
+
+ raise NotImplementedError("BPE pre-tokenizer was not recognized - update get_vocab_base_pre()")
logger.debug(f"tokenizer.ggml.pre: {repr(res)}")