Safetensors
arhansd1 commited on
Commit
f42eb70
·
verified ·
1 Parent(s): 7014e9b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -11
README.md CHANGED
@@ -11,8 +11,6 @@ base_model:
11
 
12
  **Transform English instructions with data context into executable pandas code with AI**
13
 
14
- ## Model Details
15
-
16
  ### Model Description
17
 
18
  Csv-AI-Cleaner converts **natural language instructions** into **pandas code** for data cleaning, filtering, grouping, sorting, merges, and more.
@@ -28,7 +26,6 @@ Fine-tuned on synthetic + real-world datasets using CodeT5 with LoRA for efficie
28
 
29
  - **Repository:** https://huggingface.co/arhansd1/Csv-AI-Cleaner-V3
30
 
31
- ## Uses
32
 
33
  ### Direct Use
34
  - Input: Context (sample dataset) + instruction in natural language
@@ -95,14 +92,10 @@ Instruction: Show IT department employees earning over 45000
95
  print(generate_code(input_example))
96
  ```
97
 
98
- ## Training Details
99
-
100
  ### Training Data
101
  - Combination of synthetic data cleaning instructions + public dataset column contexts
102
  - Augmented with filtered StackOverflow code snippets for pandas tasks
103
 
104
- ### Training Procedure
105
-
106
  #### Preprocessing
107
  - Normalized table format for context section
108
  - Instruction phrasing normalized to imperative form
@@ -114,7 +107,6 @@ print(generate_code(input_example))
114
  - Batch size: 8
115
  - Precision: fp16 mixed precision
116
 
117
- ## Evaluation
118
 
119
  ### Testing Data
120
  - Held-out set of 500 natural language → pandas task pairs
@@ -136,8 +128,6 @@ High syntax accuracy, good partial match rate, slightly lower exact match on mul
136
  - **Compute Region:** US-East
137
  - **Carbon Emitted:** ~1.2 kg CO2eq (estimate)
138
 
139
- ## Technical Specifications
140
-
141
  ### Model Architecture and Objective
142
  - CodeT5-base (encoder-decoder)
143
  - Objective: Seq2Seq code generation from natural language + data context
@@ -149,4 +139,4 @@ High syntax accuracy, good partial match rate, slightly lower exact match on mul
149
  ## Model Card Contact
150
  - **Author:** ArhanSD1
151
  - **Hugging Face:** https://huggingface.co/arhansd1
152
- - **Email:** [More Information Needed]
 
11
 
12
  **Transform English instructions with data context into executable pandas code with AI**
13
 
 
 
14
  ### Model Description
15
 
16
  Csv-AI-Cleaner converts **natural language instructions** into **pandas code** for data cleaning, filtering, grouping, sorting, merges, and more.
 
26
 
27
  - **Repository:** https://huggingface.co/arhansd1/Csv-AI-Cleaner-V3
28
 
 
29
 
30
  ### Direct Use
31
  - Input: Context (sample dataset) + instruction in natural language
 
92
  print(generate_code(input_example))
93
  ```
94
 
 
 
95
  ### Training Data
96
  - Combination of synthetic data cleaning instructions + public dataset column contexts
97
  - Augmented with filtered StackOverflow code snippets for pandas tasks
98
 
 
 
99
  #### Preprocessing
100
  - Normalized table format for context section
101
  - Instruction phrasing normalized to imperative form
 
107
  - Batch size: 8
108
  - Precision: fp16 mixed precision
109
 
 
110
 
111
  ### Testing Data
112
  - Held-out set of 500 natural language → pandas task pairs
 
128
  - **Compute Region:** US-East
129
  - **Carbon Emitted:** ~1.2 kg CO2eq (estimate)
130
 
 
 
131
  ### Model Architecture and Objective
132
  - CodeT5-base (encoder-decoder)
133
  - Objective: Seq2Seq code generation from natural language + data context
 
139
  ## Model Card Contact
140
  - **Author:** ArhanSD1
141
  - **Hugging Face:** https://huggingface.co/arhansd1
142
+ - **Email:** N/A