Update README.md
Browse files
README.md
CHANGED
@@ -12,14 +12,14 @@ inference: false
|
|
12 |
`{'sentiment': ['positive'], people': ['..'], 'organization': ['..'],`
|
13 |
`'place': ['..]}`
|
14 |
|
15 |
-
This 'combo' model is designed to illustrate the potential power of using function calls on small, specialized models to enable a single model architecture to combine the capabilities of what were traditionally two separate model architectures on an encoder.
|
16 |
|
17 |
The intent of SLIMs is to forge a middle-ground between traditional encoder-based classifiers and open-ended API-based LLMs, providing an intuitive, flexible natural language response, without complex prompting, and with improved generalization and ability to fine-tune to a specific domain use case.
|
18 |
|
19 |
|
20 |
This model is fine-tuned on top of [**llmware/bling-stable-lm-3b-4e1t-v0**](https://huggingface.co/llmware/bling-stable-lm-3b-4e1t-v0), which in turn, is a fine-tune of stabilityai/stablelm-3b-4elt.
|
21 |
|
22 |
-
For fast inference, we would recommend the 'quantized tool' version of this model, e.g., [**'slim-sa-ner-
|
23 |
|
24 |
|
25 |
## Prompt format:
|
@@ -33,8 +33,8 @@ For fast inference, we would recommend the 'quantized tool' version of this mode
|
|
33 |
<details>
|
34 |
<summary>Transformers Script </summary>
|
35 |
|
36 |
-
model = AutoModelForCausalLM.from_pretrained("llmware/slim-sa-ner
|
37 |
-
tokenizer = AutoTokenizer.from_pretrained("llmware/slim-sa-ner
|
38 |
|
39 |
function = "classify"
|
40 |
params = "topic"
|
|
|
12 |
`{'sentiment': ['positive'], people': ['..'], 'organization': ['..'],`
|
13 |
`'place': ['..]}`
|
14 |
|
15 |
+
This 3B parameter 'combo' model is designed to illustrate the potential power of using function calls on small, specialized models to enable a single model architecture to combine the capabilities of what were traditionally two separate model architectures on an encoder.
|
16 |
|
17 |
The intent of SLIMs is to forge a middle-ground between traditional encoder-based classifiers and open-ended API-based LLMs, providing an intuitive, flexible natural language response, without complex prompting, and with improved generalization and ability to fine-tune to a specific domain use case.
|
18 |
|
19 |
|
20 |
This model is fine-tuned on top of [**llmware/bling-stable-lm-3b-4e1t-v0**](https://huggingface.co/llmware/bling-stable-lm-3b-4e1t-v0), which in turn, is a fine-tune of stabilityai/stablelm-3b-4elt.
|
21 |
|
22 |
+
For fast inference, we would recommend the 'quantized tool' version of this model, e.g., [**'slim-sa-ner-tool'**](https://huggingface.co/llmware/slim-sa-ner-tool).
|
23 |
|
24 |
|
25 |
## Prompt format:
|
|
|
33 |
<details>
|
34 |
<summary>Transformers Script </summary>
|
35 |
|
36 |
+
model = AutoModelForCausalLM.from_pretrained("llmware/slim-sa-ner")
|
37 |
+
tokenizer = AutoTokenizer.from_pretrained("llmware/slim-sa-ner")
|
38 |
|
39 |
function = "classify"
|
40 |
params = "topic"
|