NVIDIA-NeMo / NVIDIA-NeMo/RL

Error on running SFT on the Nemotron 49B v1.5 model with NeMo-RL with mbridge backend

Open
#1,134 0 comments 0 reactions 0 assignees View on GitHub
bug t-mcore
Dominant language
Python
Stars
2k
Forks
561
Avg merge
4d 5h
Merged PRs (30d)
145

Description

Running into an error while running SFT on Nemotron Super v1.5 with NeMo-RL with mbridge backend:
Using the NeMo-RL container which uses transformers library v4.53.3.

```
✗ Could not find architecture class 'DeciLMForCausalLM' in transformers

File "/nemo-rl/nemo_rl/models/policy/megatron_policy_worker.py", line 497, in __init__
import_model_from_hf_name(
File "/nemo-rl/nemo_rl/models/megatron/community_import.py", line 35, in import_model_from_hf_name
bridge = AutoBridge.from_hf_pretrained(hf_model_name, trust_remote_code=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/nemo-rl/3rdparty/Megatron-Bridge-workspace/Megatron-Bridge/src/megatron/bridge/models/conversion/auto_bridge.py", line 228, in from_hf_pretrained
cls._validate_config(config, str(path))
File "/nemo-rl/3rdparty/Megatron-Bridge-workspace/Megatron-Bridge/src/megatron/bridge/models/conversion/auto_bridge.py", line 894, in _validate_config
raise ValueError(
ValueError:
✗ Could not find architecture class 'DeciLMForCausalLM' in transformers

This might be because:
1. The transformers library version is too old
2. The model requires a custom modeling file
3. The architecture name is incorrect

Please check your transformers installation and model requirements.

```

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.