Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
code
Size:
10K - 100K
ArXiv:
DOI:
License:
boomanaiden154
commited on
Add more dataset size information
Browse files
README.md
CHANGED
@@ -39,15 +39,37 @@ task_ids: []
|
|
39 |
|
40 |
|Release|Programming Languages|Description|
|
41 |
|-|-|-|
|
42 |
-
|v1.0| C/C++, Rust, Swift, Julia | Fine Tuning-scale dataset of
|
43 |
|
44 |
### Dataset Summary
|
45 |
|
46 |
-
ComPile contains over
|
|
|
47 |
The dataset was created by hooking into LLVM code generation either through the language's package manager or the
|
48 |
compiler directly to extract the dataset of intermediate representations from production grade programs using our
|
49 |
[dataset collection utility for the LLVM compilation infrastructure](https://doi.org/10.5281/zenodo.10155761).
|
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
### Languages
|
52 |
|
53 |
The dataset contains **5 programming languages** as of v1.0.
|
@@ -119,17 +141,17 @@ typically a link to a tar archive or git repository from which the project was b
|
|
119 |
mapping to a specific package ecosystem that provides the source, such as Spack.
|
120 |
- `language` (string): This column indicates the source language that the module was compiled from.
|
121 |
|
122 |
-
##
|
123 |
|
124 |
| Langauge | Raw Size | License Constraints | Deduplicated + License Constraints |
|
125 |
|----------|----------|---------------------|------------------------------------|
|
126 |
-
| C/C++ |
|
127 |
-
| C |
|
128 |
-
| C++ |
|
129 |
-
| Julia | 201GB | 179GB |
|
130 |
| Swift | 8GB | 7GB | 7GB |
|
131 |
-
| Rust | 656GB | 443GB |
|
132 |
-
| Total |
|
133 |
|
134 |
The raw size is the size obtained directly from building all the projects. The license constraints column
|
135 |
shows the size per language after license information is taken into account. The last column shows the size
|
@@ -138,7 +160,8 @@ dataset.
|
|
138 |
|
139 |
Note that the sizes displayed here are of the compressed bitcode representation rather
|
140 |
than textual IR. We see an expansion ratio of 2-5x, averaging around 4x when converting
|
141 |
-
from compressed bitcode to textual IR.
|
|
|
142 |
|
143 |
## Dataset Construction
|
144 |
|
|
|
39 |
|
40 |
|Release|Programming Languages|Description|
|
41 |
|-|-|-|
|
42 |
+
|v1.0| C/C++, Rust, Swift, Julia | Fine Tuning-scale dataset of 602GB of deduplicated LLVM (bitcode) IR |
|
43 |
|
44 |
### Dataset Summary
|
45 |
|
46 |
+
ComPile contains over 2.7TB of permissively-licensed source code compiled to (textual) [LLVM](https://llvm.org)
|
47 |
+
intermediate representation (IR) covering C/C++, Rust, Swift, and Julia.
|
48 |
The dataset was created by hooking into LLVM code generation either through the language's package manager or the
|
49 |
compiler directly to extract the dataset of intermediate representations from production grade programs using our
|
50 |
[dataset collection utility for the LLVM compilation infrastructure](https://doi.org/10.5281/zenodo.10155761).
|
51 |
|
52 |
+
### Dataset Size
|
53 |
+
|
54 |
+
The public release of ComPile contains over 2.7TB of textual LLVM-IR, which tokenizes into 1.3+T tokens using the Llama
|
55 |
+
tokenizer.
|
56 |
+
|
57 |
+
| Langauage | Bitcode Size | Textual IR Size | Llama Token Count | BPE Token Count (10k Vocab) | BPE Token Count (50k Vocab) |
|
58 |
+
|-----------|--------------|-----------------|-------------------|-----------------------------|-----------------------------|
|
59 |
+
| C | 2.47GB | 10.19GB | 5.31B | 0.91B | 0.58B |
|
60 |
+
| C++ | 28.87GB | 102.76GB | 46.75B | 11.20B | 6.27B |
|
61 |
+
| Julia | 164.16GB | 1088.39GB | 547.60B | 41.91B | 23.49B |
|
62 |
+
| Rust | 399.94GB | 1523.84GB | 735.90B | 137.37B | 90.01B |
|
63 |
+
| Swift | 6.95GB | 35.93GB | 19.78B | 3.36B | 1.75B |
|
64 |
+
| Total | 602.39GB | 2761.11GB | 1355.34B | 194.75B | 122.10B |
|
65 |
+
|
66 |
+
ComPile is distributed as bitcode, which is a compressed format that can be easily converted to and from the
|
67 |
+
textual representation of LLVM-IR. To collect token counts, we disassembled the bitcode to convert it into textual
|
68 |
+
IR and then ran a tokenizer over it. We used the standard Llama tokenizer and then ran fastBPE using a custom
|
69 |
+
vocabulary trained on a multi-GB sample of textual IR representativie of all languages in ComPile at two different
|
70 |
+
two different vocab sizes, particularly 10k and 50k. LLVM-IR is quite formulaic, so using custom vocabulary significantly
|
71 |
+
reduces the number of tokens generated.
|
72 |
+
|
73 |
### Languages
|
74 |
|
75 |
The dataset contains **5 programming languages** as of v1.0.
|
|
|
141 |
mapping to a specific package ecosystem that provides the source, such as Spack.
|
142 |
- `language` (string): This column indicates the source language that the module was compiled from.
|
143 |
|
144 |
+
## License Constraints and Deduplication
|
145 |
|
146 |
| Langauge | Raw Size | License Constraints | Deduplicated + License Constraints |
|
147 |
|----------|----------|---------------------|------------------------------------|
|
148 |
+
| C/C++ | 126GB | 46GB | 31GB |
|
149 |
+
| C | 16GB | N/A | 2GB |
|
150 |
+
| C++ | 109GB | N/A | 29GB |
|
151 |
+
| Julia | 201GB | 179GB | 164GB |
|
152 |
| Swift | 8GB | 7GB | 7GB |
|
153 |
+
| Rust | 656GB | 443GB | 400GB |
|
154 |
+
| Total | 990GB | 675GB | 602GB |
|
155 |
|
156 |
The raw size is the size obtained directly from building all the projects. The license constraints column
|
157 |
shows the size per language after license information is taken into account. The last column shows the size
|
|
|
160 |
|
161 |
Note that the sizes displayed here are of the compressed bitcode representation rather
|
162 |
than textual IR. We see an expansion ratio of 2-5x, averaging around 4x when converting
|
163 |
+
from compressed bitcode to textual IR. Specific per-language numbers are available in the section
|
164 |
+
above on dataset size.
|
165 |
|
166 |
## Dataset Construction
|
167 |
|