Mistral v3 (#514)

* Update llama.py

* offload

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* continued pretraining trainer

* Update trainer.py

* Update trainer.py

* Update trainer.py

* Update trainer.py

* is_bfloat16_supported

* Update __init__.py

* Update README.md

* Update llama.py

* is_bfloat16_supported

* Update __init__.py

* Mistral v3
This commit is contained in:
Daniel Han 2024-05-23 04:15:02 +10:00 committed by GitHub
commit d0b92ea58b
2 changed files with 9 additions and 0 deletions

View file

@ -17,6 +17,7 @@ import importlib
# Currently only supports 1 GPU, or else seg faults will occur.
if "CUDA_VISIBLE_DEVICES" in os.environ:
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
devices = os.environ["CUDA_VISIBLE_DEVICES"]
# Check if there are multiple cuda devices set in env
if not devices.isdigit():

View file

@ -144,6 +144,14 @@ __INT_TO_FLOAT_MAPPER = \
"unsloth/Phi-3-mini-4k-instruct",
"microsoft/Phi-3-mini-4k-instruct",
),
"unsloth/mistral-7b-v0.3-bnb-4bit" : (
"unsloth/mistral-7b-v0.3",
"mistralai/Mistral-7B-v0.3",
),
"unsloth/mistral-7b-instruct-v0.3-bnb-4bit" : (
"unsloth/mistral-7b-instruct-v0.3",
"mistralai/Mistral-7B-Instruct-v0.3",
),
}
INT_TO_FLOAT_MAPPER = {}