card: neutral what-is-in-repo wording
Browse files
README.md
CHANGED
|
@@ -91,9 +91,9 @@ def embed(texts, max_len=128, mrl_dim=768):
|
|
| 91 |
|
| 92 |
## What's in this repo — ready-to-run compiled engines
|
| 93 |
|
| 94 |
-
This repo
|
| 95 |
-
**runtime × GPU arch × OS × length-bucket**
|
| 96 |
-
hardware
|
| 97 |
|
| 98 |
- **TensorRT** `*.engine` — NVIDIA, INT8 W8A16, per arch × OS × bucket:
|
| 99 |
`code-daemon-embed-v1-{s,m,l}_{win_x64,linux_x64}_trt_sm_{86,89,120}.engine`
|
|
|
|
| 91 |
|
| 92 |
## What's in this repo — ready-to-run compiled engines
|
| 93 |
|
| 94 |
+
This repo holds **pre-compiled, ready-to-run engines**, named per
|
| 95 |
+
**runtime × GPU arch × OS × length-bucket** — grab the compiled model that matches your runtime and
|
| 96 |
+
hardware and use it directly, with no compilation on your machine.
|
| 97 |
|
| 98 |
- **TensorRT** `*.engine` — NVIDIA, INT8 W8A16, per arch × OS × bucket:
|
| 99 |
`code-daemon-embed-v1-{s,m,l}_{win_x64,linux_x64}_trt_sm_{86,89,120}.engine`
|