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
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
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
//! an rusty equivalent of `llama_token_data_array`.
use std::ptr;
use crate::error::TokenSamplingError;
use crate::sampling::LlamaSampler;
use crate::token::data::LlamaTokenData;
use super::LlamaToken;
/// a safe wrapper around `llama_token_data_array`.
#[derive(Debug, Clone, PartialEq)]
pub struct LlamaTokenDataArray {
/// the underlying data
pub data: Vec<LlamaTokenData>,
/// the index of the selected token in ``data``
pub selected: Option<usize>,
/// is the data sorted?
pub sorted: bool,
}
impl LlamaTokenDataArray {
/// Create a new `LlamaTokenDataArray` from a vector and whether or not the data is sorted.
///
/// ```
/// # use llama_cpp_bindings::token::data::LlamaTokenData;
/// # use llama_cpp_bindings::token::data_array::LlamaTokenDataArray;
/// # use llama_cpp_bindings::token::LlamaToken;
/// let array = LlamaTokenDataArray::new(vec![
/// LlamaTokenData::new(LlamaToken(0), 0.0, 0.0),
/// LlamaTokenData::new(LlamaToken(1), 0.1, 0.1)
/// ], false);
/// assert_eq!(array.data.len(), 2);
/// assert_eq!(array.sorted, false);
/// ```
#[must_use]
pub const fn new(data: Vec<LlamaTokenData>, sorted: bool) -> Self {
Self {
data,
selected: None,
sorted,
}
}
/// Create a new `LlamaTokenDataArray` from an iterator and whether or not the data is sorted.
/// ```
/// # use llama_cpp_bindings::token::data::LlamaTokenData;
/// # use llama_cpp_bindings::token::data_array::LlamaTokenDataArray;
/// # use llama_cpp_bindings::token::LlamaToken;
/// let array = LlamaTokenDataArray::from_iter([
/// LlamaTokenData::new(LlamaToken(0), 0.0, 0.0),
/// LlamaTokenData::new(LlamaToken(1), 0.1, 0.1)
/// ], false);
/// assert_eq!(array.data.len(), 2);
/// assert_eq!(array.sorted, false);
pub fn from_iter<TIterator>(data: TIterator, sorted: bool) -> Self
where
TIterator: IntoIterator<Item = LlamaTokenData>,
{
Self::new(data.into_iter().collect(), sorted)
}
/// Returns the current selected token, if one exists.
#[must_use]
pub fn selected_token(&self) -> Option<LlamaToken> {
self.data.get(self.selected?).map(LlamaTokenData::id)
}
}
impl LlamaTokenDataArray {
/// Modify the underlying data as a `llama_token_data_array`. and reconstruct the `LlamaTokenDataArray`.
///
/// # Panics
///
/// Panics if some of the safety conditions are not met. (we cannot check all of them at
/// runtime so breaking them is UB)
///
/// # Safety
///
/// The returned array formed by the data pointer and the length must entirely consist of
/// initialized token data and the length must be less than the capacity of this array's data
/// buffer.
/// If the data is not sorted, sorted must be false.
pub unsafe fn modify_as_c_llama_token_data_array<TResult>(
&mut self,
modify: impl FnOnce(&mut llama_cpp_bindings_sys::llama_token_data_array) -> TResult,
) -> TResult {
let size = self.data.len();
let data = self
.data
.as_mut_ptr()
.cast::<llama_cpp_bindings_sys::llama_token_data>();
let mut c_llama_token_data_array = llama_cpp_bindings_sys::llama_token_data_array {
data,
size,
selected: self
.selected
.and_then(|selected_index| selected_index.try_into().ok())
.unwrap_or(-1),
sorted: self.sorted,
};
let result = modify(&mut c_llama_token_data_array);
assert!(c_llama_token_data_array.size <= self.data.capacity());
// SAFETY: caller guarantees the returned data and size are valid.
unsafe {
if !ptr::eq(c_llama_token_data_array.data, data) {
ptr::copy(
c_llama_token_data_array.data,
data,
c_llama_token_data_array.size,
);
}
self.data.set_len(c_llama_token_data_array.size);
}
self.sorted = c_llama_token_data_array.sorted;
self.selected = c_llama_token_data_array
.selected
.try_into()
.ok()
.filter(|&s| s < self.data.len());
result
}
/// Modifies the data array by applying a sampler to it.
///
/// # Panics
///
/// Panics if the vendored sampler throws a C++ exception. `llama_sampler_apply` is
/// documented to be a pure logit transform and is not expected to throw; if it does
/// the failure is propagated as a panic per the crash-fast invariant.
pub fn apply_sampler(&mut self, sampler: &LlamaSampler) {
unsafe {
self.modify_as_c_llama_token_data_array(|c_llama_token_data_array| {
let mut out_error: *mut std::os::raw::c_char = ptr::null_mut();
let status = llama_cpp_bindings_sys::llama_rs_sampler_apply(
sampler.sampler,
c_llama_token_data_array,
&raw mut out_error,
);
if status != llama_cpp_bindings_sys::LLAMA_RS_SAMPLER_APPLY_OK {
let message = crate::ffi_error_reader::read_and_free_cpp_error(out_error);
panic!("llama_rs_sampler_apply returned status {status}: {message}");
}
});
}
}
/// Modifies the data array by applying a sampler to it
#[must_use]
pub fn with_sampler(mut self, sampler: &mut LlamaSampler) -> Self {
self.apply_sampler(sampler);
self
}
/// Randomly selects a token from the candidates based on their probabilities.
///
/// # Errors
/// Returns [`TokenSamplingError::NoTokenSelected`] if the sampler fails to select a token.
pub fn sample_token(&mut self, seed: u32) -> Result<LlamaToken, TokenSamplingError> {
self.apply_sampler(&LlamaSampler::dist(seed));
self.selected_token()
.ok_or(TokenSamplingError::NoTokenSelected)
}
/// Selects the token with the highest probability.
///
/// # Errors
/// Returns [`TokenSamplingError::NoTokenSelected`] if the sampler fails to select a token.
pub fn sample_token_greedy(&mut self) -> Result<LlamaToken, TokenSamplingError> {
self.apply_sampler(&LlamaSampler::greedy());
self.selected_token()
.ok_or(TokenSamplingError::NoTokenSelected)
}
}
#[cfg(test)]
mod tests {
use crate::token::LlamaToken;
use crate::token::data::LlamaTokenData;
use super::LlamaTokenDataArray;
#[test]
fn apply_greedy_sampler_selects_highest_logit() {
use crate::sampling::LlamaSampler;
let mut array = LlamaTokenDataArray::new(
vec![
LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0),
LlamaTokenData::new(LlamaToken::new(1), 5.0, 0.0),
LlamaTokenData::new(LlamaToken::new(2), 3.0, 0.0),
],
false,
);
array.apply_sampler(&LlamaSampler::greedy());
assert_eq!(array.selected_token(), Some(LlamaToken::new(1)));
}
#[test]
fn with_sampler_builder_pattern() {
use crate::sampling::LlamaSampler;
let array = LlamaTokenDataArray::new(
vec![
LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0),
LlamaTokenData::new(LlamaToken::new(1), 5.0, 0.0),
],
false,
)
.with_sampler(&mut LlamaSampler::greedy());
assert_eq!(array.selected_token(), Some(LlamaToken::new(1)));
}
#[test]
fn sample_token_greedy_returns_highest() {
let mut array = LlamaTokenDataArray::new(
vec![
LlamaTokenData::new(LlamaToken::new(10), 0.1, 0.0),
LlamaTokenData::new(LlamaToken::new(20), 9.9, 0.0),
],
false,
);
let token = array
.sample_token_greedy()
.expect("test: greedy sampler should select a token");
assert_eq!(token, LlamaToken::new(20));
}
#[test]
fn from_iter_creates_array_from_iterator() {
let array = LlamaTokenDataArray::from_iter(
[
LlamaTokenData::new(LlamaToken::new(0), 0.0, 0.0),
LlamaTokenData::new(LlamaToken::new(1), 1.0, 0.0),
LlamaTokenData::new(LlamaToken::new(2), 2.0, 0.0),
],
false,
);
assert_eq!(array.data.len(), 3);
assert!(!array.sorted);
assert!(array.selected.is_none());
}
#[test]
fn sample_token_with_seed_selects_a_token() {
let mut array = LlamaTokenDataArray::new(
vec![
LlamaTokenData::new(LlamaToken::new(10), 1.0, 0.0),
LlamaTokenData::new(LlamaToken::new(20), 1.0, 0.0),
],
false,
);
let token = array
.sample_token(42)
.expect("test: dist sampler should select a token");
assert!(token == LlamaToken::new(10) || token == LlamaToken::new(20));
}
#[test]
fn selected_token_returns_none_when_no_selection() {
let array = LlamaTokenDataArray::new(
vec![LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0)],
false,
);
assert!(array.selected_token().is_none());
}
#[test]
fn selected_token_returns_none_when_index_out_of_bounds() {
let array = LlamaTokenDataArray {
data: vec![LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0)],
selected: Some(5),
sorted: false,
};
assert!(array.selected_token().is_none());
}
#[test]
fn modify_as_c_llama_token_data_array_copies_when_data_pointer_changes() {
let mut array = LlamaTokenDataArray::new(
vec![
LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0),
LlamaTokenData::new(LlamaToken::new(1), 2.0, 0.0),
LlamaTokenData::new(LlamaToken::new(2), 3.0, 0.0),
],
false,
);
let replacement = [
llama_cpp_bindings_sys::llama_token_data {
id: 10,
logit: 5.0,
p: 0.0,
},
llama_cpp_bindings_sys::llama_token_data {
id: 20,
logit: 6.0,
p: 0.0,
},
];
unsafe {
array.modify_as_c_llama_token_data_array(|c_array| {
c_array.data = replacement.as_ptr().cast_mut();
c_array.size = replacement.len();
c_array.selected = 0;
});
}
assert_eq!(array.data.len(), 2);
assert_eq!(array.data[0].id(), LlamaToken::new(10));
assert_eq!(array.data[1].id(), LlamaToken::new(20));
assert_eq!(array.selected, Some(0));
}
#[test]
fn selected_overflow_uses_negative_one() {
let mut array = LlamaTokenDataArray {
data: vec![LlamaTokenData::new(LlamaToken::new(0), 1.0, 0.0)],
selected: Some(usize::MAX),
sorted: false,
};
unsafe {
array.modify_as_c_llama_token_data_array(|c_array| {
assert_eq!(c_array.selected, -1);
});
}
}
}