Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,42 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
---
|
4 |
+
### environment
|
5 |
+
optimum-neuron 0.0.25
|
6 |
+
|
7 |
+
neuron 2.20.0
|
8 |
+
|
9 |
+
transformers-neuronx 0.12.313
|
10 |
+
|
11 |
+
transformers 4.45.2
|
12 |
+
|
13 |
+
|
14 |
+
### export
|
15 |
+
```
|
16 |
+
optimum-cli export neuron --model meta-llama/Llama-3.2-1B-Instruct --batch_size 1 --sequence_length 1024 --num_cores 2 --auto_cast_type fp16 ./models-hf/meta-llama/Llama-3.2-1B-Instruct
|
17 |
+
|
18 |
+
```
|
19 |
+
|
20 |
+
### run
|
21 |
+
```
|
22 |
+
docker run -it --name llama-31 --rm \
|
23 |
+
-p 8080:80 \
|
24 |
+
-v /home/ec2-user/models-hf/:/models \
|
25 |
+
-e HF_MODEL_ID=/models/meta-llama/Llama-3.2-1B-Instruct \
|
26 |
+
-e MAX_INPUT_TOKENS=256 \
|
27 |
+
-e MAX_TOTAL_TOKENS=4096 \
|
28 |
+
-e MAX_BATCH_SIZE=1 \
|
29 |
+
-e LOG_LEVEL="info,text_generation_router=debug,text_generation_launcher=debug" \
|
30 |
+
--device=/dev/neuron0 \
|
31 |
+
neuronx-tgi:latest \
|
32 |
+
--model-id /models/meta-llama/Llama-3.2-1B-Instruct \
|
33 |
+
--max-batch-size 1 \
|
34 |
+
--max-input-tokens 256 \
|
35 |
+
--max-total-tokens 1024
|
36 |
+
|
37 |
+
```
|
38 |
+
|
39 |
+
### test
|
40 |
+
```
|
41 |
+
curl 127.0.0.1:8080/generate -X POST -d '{"inputs":"What is Deep Learning?","parameters":{"max_new_tokens":20}}' -H 'Content-Type: application/json'
|
42 |
+
```
|