-
### 🚀 The feature, motivation and pitch
With a single command, quantize the same model across every available quant scheme and configuration and output a table that compares the results. This will …
byjlw updated
1 month ago
-
While running HF Llama 3.2 1b with 1 layer in Thunder under the PyTorch profiler, we noticed that apparently, the NVFuser's run_fused_kernel takes a lot of time.
https://github.com/Lightning-AI/lightn…
-
I ran HuggingFace BERT model which uses tensorflow 2.13v with oneDNN support on intel machine and recorded its execution logs by setting TF_CPP_MAX_VLOG_LEVEL=2 & ONEDNN_VERBOSE=1 in file.
**Observ…
-
| --- | --- |
| Bugzilla Link | [516652](https://bugs.eclipse.org/bugs/show_bug.cgi?id=516652) |
| Status | NEW |
| Importance | P3 normal |
| Reported | May 15, 2017 09:51 EDT |
| Modified | Jul…
-
Update: these are the issues we're tracking for trace e2e perf
- https://github.com/tenstorrent/tt-metal/issues/11398
- https://github.com/tenstorrent/tt-metal/issues/11509
Running Llama deco…
-
### Description
I'm trying to restore an RLLib algorithm from a checkpoint and change the configuration before resuming training. My main objective is to change the number of rollout workers between …
-
I installed all the dependencies and ran the code , I have this error
File "/home/lnuser/anaconda3/lib/python3.6/site-packages/tensorflow/python/ops/gradients_util.py", line 504, in _GradientsHelpe…
-
When I try to scale up my model, I find that `tf.scan` uses too much memory. Probably I have to use eager execution.
How well does `HMC` and `SGMCMC` work with eager execution and tensorflow 2.0?
-
The following test aims to make sure that we walk the array from right to left and checks it in the following way: multiplying left to right 1e-40 cannot be held into a float32 so it causes an underfl…
-
### Describe the bug
In eager workflows, if the future/awaitable of a Flyte-task that returns a tuple is wrapped in an asyncio task (Python), then the result of that task will be a named tuple when e…