Giters
mosaicml
/
examples
Fast and flexible reference benchmarks
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
419
Watchers:
16
Issues:
37
Forks:
120
mosaicml/examples Issues
How to add a custom key to config file?
Updated
2 months ago
Error when training with Mosaic-Bert
Updated
4 months ago
MosaicBERT: Convert composer weights to HF
Updated
4 months ago
Comments count
1
MosaicBERT: pretraining configuration for models > 128 seq. length
Updated
5 months ago
Comments count
5
Change bf16 to amp_bf16
Updated
5 months ago
Comments count
3
FlashAttention Triton error on the MosaicBERT models other than base
Closed
5 months ago
Comments count
3
config class for bert is not consistent
Updated
5 months ago
Comments count
2
Please bring code features from MPT-7b back to MPT-1b for use of MPT-1b with SFTTrainer.
Updated
6 months ago
RuntimeError: Triton Error [CUDA]: invalid argument
Updated
8 months ago
Comments count
17
Can't save a trained model as a HuggingFace model
Closed
10 months ago
Comments count
5
CUDA out of memory
Closed
10 months ago
Comments count
2
--concat_tokens flag in BERT pretraining
Closed
10 months ago
Comments count
2
Pre-commit checks
Closed
10 months ago
Comments count
1
Update Readme
Closed
10 months ago
Comments count
1
Regression testing
Closed
10 months ago
Comments count
1
1 out of N runs starts successfully, others fail immediately
Closed
10 months ago
Comments count
9
Accessing model after pre-training
Closed
10 months ago
Comments count
1
Inquiry about Mosaic-BERT and BERT-Base Sequence Lengths
Closed
a year ago
Comments count
9
Finetuning script broken?
Closed
10 months ago
Comments count
4
Finetuning on windows machine
Closed
a year ago
Comments count
4
Confusion regarding conflicting information in model card of "mosaic-bert" on Hugging Face
Closed
a year ago
Comments count
2
Explain composer logs emitted during training + Replicate Benchmark Results
Closed
a year ago
Comments count
1
Train BERT on own data
Closed
a year ago
Comments count
3
Training Time estimation on single GPU A100 80G
Closed
a year ago
Comments count
9
Error using PIL
Closed
a year ago
Comments count
3
link to bert example is broken
Closed
a year ago
Comments count
2
will fine tuning work on windows 11 lap top minimal gpu for bert sequence_classification.py ?
Closed
a year ago
Comments count
5
Integration documentation is broken
Updated
a year ago
Comments count
1
FSDP for encoder
Closed
a year ago
ValueError: Value bf16 not found in Precision
Closed
a year ago
[Question] Use flash attention w/ RedPajama LLM?
Closed
a year ago
Loss spike when training mosaic-bert (fp32)
Closed
a year ago
Comments count
3
Error in ResNet ImageNet examples
Closed
a year ago
Comments count
1
Default num_canonical_nodes to an even multiple of num_physical_nodes
Updated
a year ago
Comments count
3
MosaicML LLM: 'key_padding_mask' is NoneType when setting "attn_impl: torch"
Closed
a year ago
Comments count
2
Confusing comment in the deeplabv3.yaml
Updated
a year ago
Comments count
1
Remove ComposerClassifier from vision benchmarks
Updated
2 years ago