Update README.md
Browse files
README.md
CHANGED
@@ -1,92 +1,75 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
The
|
41 |
-
|
42 |
-
-
|
43 |
-
|
44 |
-
-
|
45 |
-
|
46 |
-
-
|
47 |
-
|
48 |
-
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
|
70 |
-
|
71 |
-
|
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
| 0.2158 | 5.81 | 2300 | 0.6875 | 0.7175 | 0.7266 |
|
77 |
-
| 0.2158 | 6.06 | 2400 | 0.6544 | 0.7236 | 0.7296 |
|
78 |
-
| 0.1423 | 6.31 | 2500 | 0.6738 | 0.7236 | 0.7313 |
|
79 |
-
| 0.1423 | 6.57 | 2600 | 0.6640 | 0.7175 | 0.7253 |
|
80 |
-
| 0.1423 | 6.82 | 2700 | 0.6617 | 0.7154 | 0.7233 |
|
81 |
-
| 0.1423 | 7.07 | 2800 | 0.6582 | 0.7154 | 0.7205 |
|
82 |
-
| 0.1423 | 7.32 | 2900 | 0.6678 | 0.7033 | 0.7093 |
|
83 |
-
| 0.1204 | 7.58 | 3000 | 0.6596 | 0.7154 | 0.7197 |
|
84 |
-
| 0.1204 | 7.83 | 3100 | 0.6598 | 0.7154 | 0.7217 |
|
85 |
-
|
86 |
-
|
87 |
-
### Framework versions
|
88 |
-
|
89 |
-
- Transformers 4.37.0
|
90 |
-
- Pytorch 2.1.2
|
91 |
-
- Datasets 2.1.0
|
92 |
-
- Tokenizers 0.15.1
|
|
|
1 |
+
# User Flow Text Classification
|
2 |
+
|
3 |
+
This model is a fined-tuned version of [nreimers/MiniLMv2-L6-H384-distilled-from-RoBERTa-Large](https://huggingface.co/nreimers/MiniLMv2-L6-H384-distilled-from-RoBERTa-Large).
|
4 |
+
The quantized version in ONNX format can be found [here](https://huggingface.co/minuva/MiniLMv2-userflow-v2-onnx)
|
5 |
+
|
6 |
+
A flow label is orthogonal to the main conversation goal, implying that it categorizes actions or responses in a way that is independent from the primary objective of the conversation.
|
7 |
+
|
8 |
+
# Load the Model
|
9 |
+
|
10 |
+
```py
|
11 |
+
from transformers import pipeline
|
12 |
+
|
13 |
+
pipe = pipeline(model='minuva/MiniLMv2-userflow-v2', task='text-classification')
|
14 |
+
pipe("This is wrong")
|
15 |
+
# [{'label': 'model_wrong_or_try_again', 'score': 0.9729849100112915}]
|
16 |
+
```
|
17 |
+
# Categories Explanation
|
18 |
+
|
19 |
+
<details>
|
20 |
+
<summary>Click to expand!</summary>
|
21 |
+
|
22 |
+
- OTHER: Responses that do not fit into any predefined categories or are outside the scope of the specific interaction types listed.
|
23 |
+
|
24 |
+
- agrees_praising_thanking: When the user agrees with the provided information, offers praise, or expresses gratitude.
|
25 |
+
|
26 |
+
- asks_source: The user requests the source of the information or the basis for the answer provided.
|
27 |
+
|
28 |
+
- continue: Indicates a prompt for the conversation to proceed or continue without a specific directional change.
|
29 |
+
|
30 |
+
- continue_or_finnish_code: Signals either to continue with the current line of discussion or code execution, or to conclude it.
|
31 |
+
|
32 |
+
- improve_or_modify_answer: The user requests an improvement or modification to the provided answer.
|
33 |
+
|
34 |
+
- lack_of_understandment: Reflects the user's or agent confusion or lack of understanding regarding the information provided.
|
35 |
+
|
36 |
+
- model_wrong_or_try_again: Indicates that the model's response was incorrect or unsatisfactory, suggesting a need to attempt another answer.
|
37 |
+
|
38 |
+
- more_listing_or_expand: The user requests further elaboration, expansion from the given list by the agent.
|
39 |
+
|
40 |
+
- repeat_answers_or_question: The need to reiterate a previous answer or question.
|
41 |
+
|
42 |
+
- request_example: The user asks for examples to better understand the concept or answer provided.
|
43 |
+
|
44 |
+
- user_complains_repetition: The user notes that the information or responses are repetitive, indicating a need for new or different content.
|
45 |
+
|
46 |
+
- user_doubts_answer: The user expresses skepticism or doubt regarding the accuracy or validity of the provided answer.
|
47 |
+
|
48 |
+
- user_goodbye: The user says goodbye to the agent.
|
49 |
+
|
50 |
+
- user_reminds_question: The user reiterates the question.
|
51 |
+
|
52 |
+
- user_wants_agent_to_answer: The user explicitly requests a response from the agent, when the agent refuses to do so.
|
53 |
+
|
54 |
+
- user_wants_explanation: The user seeks an explanation behind the information or answer provided.
|
55 |
+
|
56 |
+
- user_wants_more_detail: Indicates the user's desire for more comprehensive or detailed information on the topic.
|
57 |
+
|
58 |
+
- user_wants_shorter_longer_answer: The user requests that the answer be condensed or expanded to better meet their informational needs.
|
59 |
+
|
60 |
+
- user_wants_simplier_explanation: The user seeks a simpler, more easily understood explanation.
|
61 |
+
|
62 |
+
- user_wants_yes_or_no: The user is asking for a straightforward affirmative or negative answer, without additional detail or explanation.
|
63 |
+
</details>
|
64 |
+
|
65 |
+
<br>
|
66 |
+
|
67 |
+
|
68 |
+
# Metrics in our private test dataset
|
69 |
+
| Model (params) | Loss | Accuracy | F1 |
|
70 |
+
|--------------------|-------------|----------|--------|
|
71 |
+
| minuva/MiniLMv2-userflow-v2 (33M) | 0.6738 | 0.7236 | 0.7313 |
|
72 |
+
|
73 |
+
# Deployment
|
74 |
+
|
75 |
+
Check [our repository](https://github.com/minuva/flow-cloudrun) to see how to easily deploy this (quantized) model in a serverless environment with fast CPU inference and light resource utilization.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|