mradermacher

/

model_requests

like 126

mradermacher was faster and already queued it as highly anticipated model.

You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#SimpleChat-30BA3B-V3-GGUF for quants to appear.

["https://huggingface.co/OpenBuddy/SimpleChat-30BA3B-V3",["worker","+cork","s","i","0"],1757910436],
https://huggingface.co/OpenBuddy/SimpleChat-30BA3B-V3 already in llmjob.submit.txt

Turns out this model failed in the past with the following error:

SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.ffn_gate_inp.weight,  torch.bfloat16 --> F32, shape = {2048, 128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_k_norm.weight,   torch.bfloat16 --> F32, shape = {128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_k.weight,        torch.bfloat16 --> BF16, shape = {2048, 512}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_output.weight,   torch.bfloat16 --> BF16, shape = {4096, 2048}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_q_norm.weight,   torch.bfloat16 --> F32, shape = {128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_q.weight,        torch.bfloat16 --> BF16, shape = {2048, 4096}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_v.weight,        torch.bfloat16 --> BF16, shape = {2048, 512}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: loading model part 'model-00013-of-00013.safetensors'
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:output.weight,               torch.bfloat16 --> BF16, shape = {2048, 151936}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.attn_norm.weight,     torch.bfloat16 --> F32, shape = {2048}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {768, 2048, 128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {2048, 768, 128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.ffn_up_exps.weight,   torch.bfloat16 --> BF16, shape = {2048, 768, 128}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:blk.47.ffn_norm.weight,      torch.bfloat16 --> F32, shape = {2048}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:output_norm.weight,          torch.bfloat16 --> F32, shape = {2048}
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:Set meta model
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:Set model parameters
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: context length = 200000
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: embedding length = 2048
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: feed forward length = 5472
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: head count = 32
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: key-value head count = 4
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: rope theta = 1000000
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: rms norm epsilon = 1e-06
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: experts used count = 8
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: file type = 1025
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: expert feed forward length = 768
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:gguf: expert shared feed forward length = 0
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:Set model quantization version
SimpleChat-30BA3B-V3    INFO:hf-to-gguf:Set model tokenizer
SimpleChat-30BA3B-V3    Traceback (most recent call last):
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 9002, in <module>
SimpleChat-30BA3B-V3        main()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 8996, in main
SimpleChat-30BA3B-V3        model_instance.write()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 440, in write
SimpleChat-30BA3B-V3        self.prepare_metadata(vocab_only=False)
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 561, in prepare_metadata
SimpleChat-30BA3B-V3        self.set_vocab()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 3694, in set_vocab
SimpleChat-30BA3B-V3        super().set_vocab()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 533, in set_vocab
SimpleChat-30BA3B-V3        self._set_vocab_gpt2()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 924, in _set_vocab_gpt2
SimpleChat-30BA3B-V3        tokens, toktypes, tokpre = self.get_vocab_base()
SimpleChat-30BA3B-V3      File "/llmjob/llama.cpp/convert_hf_to_gguf.py", line 647, in get_vocab_base
SimpleChat-30BA3B-V3        tokenizer = AutoTokenizer.from_pretrained(self.dir_model)
SimpleChat-30BA3B-V3      File "/llmjob/share/python/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 1135, in from_pretrained
SimpleChat-30BA3B-V3        return tokenizer_class_fast.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
SimpleChat-30BA3B-V3      File "/llmjob/share/python/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2069, in from_pretrained
SimpleChat-30BA3B-V3        return cls._from_pretrained(
SimpleChat-30BA3B-V3      File "/llmjob/share/python/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2107, in _from_pretrained
SimpleChat-30BA3B-V3        slow_tokenizer = (cls.slow_tokenizer_class)._from_pretrained(
SimpleChat-30BA3B-V3      File "/llmjob/share/python/lib/python3.10/site-packages/transformers/tokenization_utils_base.py", line 2315, in _from_pretrained
SimpleChat-30BA3B-V3        tokenizer = cls(*init_inputs, **init_kwargs)
SimpleChat-30BA3B-V3      File "/llmjob/share/python/lib/python3.10/site-packages/transformers/models/qwen2/tokenization_qwen2.py", line 172, in __init__
SimpleChat-30BA3B-V3        with open(vocab_file, encoding="utf-8") as vocab_handle:
SimpleChat-30BA3B-V3    TypeError: expected str, bytes or os.PathLike object, not NoneType
SimpleChat-30BA3B-V3    yes: standard output: Broken pipe
SimpleChat-30BA3B-V3    job finished, status 1
SimpleChat-30BA3B-V3    job-done<0 SimpleChat-30BA3B-V3 noquant 1>
SimpleChat-30BA3B-V3
SimpleChat-30BA3B-V3    NAME: SimpleChat-30BA3B-V3
SimpleChat-30BA3B-V3    TIME: Mon Sep 15 09:59:20 2025
SimpleChat-30BA3B-V3    WORKER: nico1

https://huggingface.co/OpenBuddy/SimpleChat-30BA3B-V3/discussions/1

According to model authors, "...the tokenizer file of this model has be failed for upload. Please copy the tokenizer files from OpenBuddy/SimpleChat-30BA3B-V2 and try again." I tried this locally, and the conversion is working. You need to obtain all those JSON files from from OpenBuddy/SimpleChat-30BA3B-V2.

Great. I will try to do so. I will let you know shortly if it works.

It's queued! :D
It successfully manually converted after following the model author’s instructions and downloaded all the required files from OpenBuddy/SimpleChat-30BA3B-V2.
The model is currently getting progressed by nico1. Weighted/imatrix quants will be delayed by a few hours because of the currently ongoing shisa-v2-llama3.1-405b RPC imatrix computation.

You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#SimpleChat-30BA3B-V3-GGUF for quants to appear.