-
hi, for the txt image genreateion, why not try using existed LLMs, plus the decoder in tokenizer, and training the whole model with lora?
just like SEEd does.
So that it at least won't harm the …
-
Is the argument to `embed` expected to be a monad morphism? I'm *guessing* so? I believe that having only a `Monad n` constraint, and not a `Monad m` one, makes it impossible to write sensible instanc…
-
Hi,
How to fine-tune MMS TTS models. I used the default vits code, however, i had issues when resuming from the existing optimizer state dict:
"
in adamw
exp_avg.mul_(beta1).add_(grad, alpha…
-
## 🚀 Feature
Remove the redundant `num_nodes` Trainer argument. Knowing the number of nodes is not required, and the world size is provided by the cluster environment anyway.
### Motivation
…
-
### Model Series
Qwen2
### What are the models used?
Qwen2-57B-A14B
### What is the scenario where the problem happened?
train with transformers
### Is this a known issue?
- [X] I have followed…
FL77N updated
2 weeks ago
-
I do a fresh installation of the latest h2ogpt. When i hit the following command
`python generate.py --base_model=meta-llama/Llama-2-7b-chat-hf --score_model=None --langchain_mode='UserData' --use…
-
## 🚀 Feature
### Motivation
After reading through PyTorch codebase, I came across this code example:
https://github.com/pytorch/pytorch/blob/538647fe1fb94b7822ea3b8bbbd6901961431d60/torch/f…
-
Is there for example an instance:
```haskell
instance Selective f => Selective (ReaderT r f a)
```
-
## 🚀 Feature
Provide Join through an intuitive API in LightningLite and make it backend agnostic, i.e., switching from DDP to single-device and vice versa should not require changes to the code.
…
-
## 🚀 Feature
Introducing the support for `precision="auto"`, which enables AMP whenever available and falls back to fp32 otherwise.
### Motivation
After introducing `accelerator="auto"` s…