Create README_TEMPLATE.md
Browse files- README_TEMPLATE.md +62 -0
README_TEMPLATE.md
ADDED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- llm-rs
|
4 |
+
- ggml
|
5 |
+
pipeline_tag: text-generation
|
6 |
+
license: apache-2.0
|
7 |
+
language:
|
8 |
+
- en
|
9 |
+
---
|
10 |
+
|
11 |
+
# GGML converted versions of [EleutherAI](https://huggingface.co/EleutherAI)'s Pythia models
|
12 |
+
|
13 |
+
## Description:
|
14 |
+
The *Pythia Scaling Suite* is a collection of models developed to facilitate
|
15 |
+
interpretability research. It contains two sets of eight models of sizes
|
16 |
+
70M, 160M, 410M, 1B, 1.4B, 2.8B, 6.9B, and 12B. For each size, there are two
|
17 |
+
models: one trained on the Pile, and one trained on the Pile after the dataset
|
18 |
+
has been globally deduplicated. All 8 model sizes are trained on the exact
|
19 |
+
same data, in the exact same order. We also provide 154 intermediate
|
20 |
+
checkpoints per model, hosted on Hugging Face as branches.
|
21 |
+
|
22 |
+
The Pythia model suite was deliberately designed to promote scientific
|
23 |
+
research on large language models, especially interpretability research.
|
24 |
+
Despite not centering downstream performance as a design goal, we find the
|
25 |
+
models match or exceed the performance of
|
26 |
+
similar and same-sized models, such as those in the OPT and GPT-Neo suites.
|
27 |
+
|
28 |
+
## Converted Models:
|
29 |
+
|
30 |
+
$MODELS$
|
31 |
+
|
32 |
+
## Usage
|
33 |
+
|
34 |
+
### Python via [llm-rs](https://github.com/LLukas22/llm-rs-python):
|
35 |
+
|
36 |
+
#### Installation
|
37 |
+
Via pip: `pip install llm-rs`
|
38 |
+
|
39 |
+
#### Run inference
|
40 |
+
```python
|
41 |
+
from llm_rs import AutoModel
|
42 |
+
|
43 |
+
#Load the model, define any model you like from the list above as the `model_file`
|
44 |
+
model = AutoModel.from_pretrained("rustformers/pythia-ggml",model_file="pythia-70m-q4_0-ggjt.bin")
|
45 |
+
|
46 |
+
#Generate
|
47 |
+
print(model.generate("The meaning of life is"))
|
48 |
+
```
|
49 |
+
|
50 |
+
### Rust via [Rustformers/llm](https://github.com/rustformers/llm):
|
51 |
+
|
52 |
+
#### Installation
|
53 |
+
```
|
54 |
+
git clone --recurse-submodules https://github.com/rustformers/llm.git
|
55 |
+
cd llm
|
56 |
+
cargo build --release
|
57 |
+
```
|
58 |
+
|
59 |
+
#### Run inference
|
60 |
+
```
|
61 |
+
cargo run --release -- gptneox infer -m path/to/model.bin -p "Tell me how cool the Rust programming language is:"
|
62 |
+
```
|