|
| 1 | +// gen_inference_defaults fetches unsloth's inference_defaults.json, |
| 2 | +// validates its structure, remaps field names to LocalAI conventions, |
| 3 | +// and writes the result to core/config/inference_defaults.json. |
| 4 | +// |
| 5 | +// Run via: go generate ./core/config/ |
| 6 | +package main |
| 7 | + |
| 8 | +import ( |
| 9 | + "encoding/json" |
| 10 | + "fmt" |
| 11 | + "io" |
| 12 | + "net/http" |
| 13 | + "os" |
| 14 | + "sort" |
| 15 | + "strings" |
| 16 | +) |
| 17 | + |
| 18 | +const ( |
| 19 | + unslothURL = "https://raw.githubusercontent.com/unslothai/unsloth/main/studio/backend/assets/configs/inference_defaults.json" |
| 20 | + outputFile = "inference_defaults.json" |
| 21 | +) |
| 22 | + |
| 23 | +// unslothDefaults mirrors the upstream JSON structure |
| 24 | +type unslothDefaults struct { |
| 25 | + Comment string `json:"_comment"` |
| 26 | + Families map[string]map[string]float64 `json:"families"` |
| 27 | + Patterns []string `json:"patterns"` |
| 28 | +} |
| 29 | + |
| 30 | +// localAIDefaults is our output structure |
| 31 | +type localAIDefaults struct { |
| 32 | + Comment string `json:"_comment"` |
| 33 | + Families map[string]map[string]float64 `json:"families"` |
| 34 | + Patterns []string `json:"patterns"` |
| 35 | +} |
| 36 | + |
| 37 | +// requiredFields are the fields every family entry must have |
| 38 | +var requiredFields = []string{"temperature", "top_p", "top_k"} |
| 39 | + |
| 40 | +// fieldRemap maps unsloth field names to LocalAI field names |
| 41 | +var fieldRemap = map[string]string{ |
| 42 | + "repetition_penalty": "repeat_penalty", |
| 43 | +} |
| 44 | + |
| 45 | +// allowedFields are the only fields we keep (after remapping) |
| 46 | +var allowedFields = map[string]bool{ |
| 47 | + "temperature": true, |
| 48 | + "top_p": true, |
| 49 | + "top_k": true, |
| 50 | + "min_p": true, |
| 51 | + "repeat_penalty": true, |
| 52 | + "presence_penalty": true, |
| 53 | +} |
| 54 | + |
| 55 | +func main() { |
| 56 | + fmt.Fprintf(os.Stderr, "Fetching %s ...\n", unslothURL) |
| 57 | + |
| 58 | + resp, err := http.Get(unslothURL) |
| 59 | + if err != nil { |
| 60 | + fatal("fetch failed: %v", err) |
| 61 | + } |
| 62 | + defer resp.Body.Close() |
| 63 | + |
| 64 | + if resp.StatusCode != 200 { |
| 65 | + fatal("fetch returned HTTP %d", resp.StatusCode) |
| 66 | + } |
| 67 | + |
| 68 | + body, err := io.ReadAll(resp.Body) |
| 69 | + if err != nil { |
| 70 | + fatal("read body: %v", err) |
| 71 | + } |
| 72 | + |
| 73 | + var upstream unslothDefaults |
| 74 | + if err := json.Unmarshal(body, &upstream); err != nil { |
| 75 | + fatal("parse upstream JSON: %v", err) |
| 76 | + } |
| 77 | + |
| 78 | + // Validate structure |
| 79 | + if len(upstream.Families) == 0 { |
| 80 | + fatal("upstream has no families") |
| 81 | + } |
| 82 | + if len(upstream.Patterns) == 0 { |
| 83 | + fatal("upstream has no patterns") |
| 84 | + } |
| 85 | + |
| 86 | + // Validate every pattern references a family |
| 87 | + for _, p := range upstream.Patterns { |
| 88 | + if _, ok := upstream.Families[p]; !ok { |
| 89 | + fatal("pattern %q has no corresponding family entry", p) |
| 90 | + } |
| 91 | + } |
| 92 | + |
| 93 | + // Validate every family has required fields and remap field names |
| 94 | + output := localAIDefaults{ |
| 95 | + Comment: "Auto-generated from unsloth inference_defaults.json. DO NOT EDIT. Run go generate ./core/config/ to update.", |
| 96 | + Families: make(map[string]map[string]float64, len(upstream.Families)), |
| 97 | + Patterns: upstream.Patterns, |
| 98 | + } |
| 99 | + |
| 100 | + // Sort family names for deterministic output |
| 101 | + familyNames := make([]string, 0, len(upstream.Families)) |
| 102 | + for name := range upstream.Families { |
| 103 | + familyNames = append(familyNames, name) |
| 104 | + } |
| 105 | + sort.Strings(familyNames) |
| 106 | + |
| 107 | + for _, name := range familyNames { |
| 108 | + params := upstream.Families[name] |
| 109 | + |
| 110 | + // Check required fields |
| 111 | + for _, req := range requiredFields { |
| 112 | + found := false |
| 113 | + for k := range params { |
| 114 | + mapped := k |
| 115 | + if m, ok := fieldRemap[k]; ok { |
| 116 | + mapped = m |
| 117 | + } |
| 118 | + if mapped == req || k == req { |
| 119 | + found = true |
| 120 | + break |
| 121 | + } |
| 122 | + } |
| 123 | + if !found { |
| 124 | + fatal("family %q missing required field %q", name, req) |
| 125 | + } |
| 126 | + } |
| 127 | + |
| 128 | + // Remap and filter fields |
| 129 | + remapped := make(map[string]float64) |
| 130 | + for k, v := range params { |
| 131 | + if newName, ok := fieldRemap[k]; ok { |
| 132 | + k = newName |
| 133 | + } |
| 134 | + if allowedFields[k] { |
| 135 | + remapped[k] = v |
| 136 | + } |
| 137 | + } |
| 138 | + output.Families[name] = remapped |
| 139 | + } |
| 140 | + |
| 141 | + // Validate patterns are ordered longest-match-first within same prefix groups |
| 142 | + validatePatternOrder(output.Patterns) |
| 143 | + |
| 144 | + // Marshal with ordered keys for readability |
| 145 | + data, err := marshalOrdered(output) |
| 146 | + if err != nil { |
| 147 | + fatal("marshal output: %v", err) |
| 148 | + } |
| 149 | + |
| 150 | + if err := os.WriteFile(outputFile, data, 0644); err != nil { |
| 151 | + fatal("write %s: %v", outputFile, err) |
| 152 | + } |
| 153 | + |
| 154 | + fmt.Fprintf(os.Stderr, "Written %s (%d families, %d patterns)\n", |
| 155 | + outputFile, len(output.Families), len(output.Patterns)) |
| 156 | +} |
| 157 | + |
| 158 | +// validatePatternOrder warns if a shorter pattern appears before a longer one |
| 159 | +// that it's a prefix of (e.g., "qwen3" before "qwen3.5") |
| 160 | +func validatePatternOrder(patterns []string) { |
| 161 | + for i, p := range patterns { |
| 162 | + for j := i + 1; j < len(patterns); j++ { |
| 163 | + if strings.HasPrefix(patterns[j], p) { |
| 164 | + fmt.Fprintf(os.Stderr, "WARNING: pattern %q at index %d is a prefix of %q at index %d — longer match should come first\n", |
| 165 | + p, i, patterns[j], j) |
| 166 | + } |
| 167 | + } |
| 168 | + } |
| 169 | +} |
| 170 | + |
| 171 | +// marshalOrdered produces JSON with families in pattern order for readability |
| 172 | +func marshalOrdered(d localAIDefaults) ([]byte, error) { |
| 173 | + var sb strings.Builder |
| 174 | + sb.WriteString("{\n") |
| 175 | + sb.WriteString(fmt.Sprintf(" %q: %q,\n", "_comment", d.Comment)) |
| 176 | + sb.WriteString(" \"families\": {\n") |
| 177 | + |
| 178 | + // Write families in pattern order, then any remaining not in patterns |
| 179 | + written := make(map[string]bool) |
| 180 | + allFamilies := make([]string, 0, len(d.Families)) |
| 181 | + for _, p := range d.Patterns { |
| 182 | + if _, ok := d.Families[p]; ok && !written[p] { |
| 183 | + allFamilies = append(allFamilies, p) |
| 184 | + written[p] = true |
| 185 | + } |
| 186 | + } |
| 187 | + for name := range d.Families { |
| 188 | + if !written[name] { |
| 189 | + allFamilies = append(allFamilies, name) |
| 190 | + } |
| 191 | + } |
| 192 | + |
| 193 | + for i, name := range allFamilies { |
| 194 | + params := d.Families[name] |
| 195 | + paramJSON, err := json.Marshal(params) |
| 196 | + if err != nil { |
| 197 | + return nil, err |
| 198 | + } |
| 199 | + comma := "," |
| 200 | + if i == len(allFamilies)-1 { |
| 201 | + comma = "" |
| 202 | + } |
| 203 | + sb.WriteString(fmt.Sprintf(" %q: %s%s\n", name, paramJSON, comma)) |
| 204 | + } |
| 205 | + |
| 206 | + sb.WriteString(" },\n") |
| 207 | + |
| 208 | + // Patterns array |
| 209 | + patternsJSON, err := json.Marshal(d.Patterns) |
| 210 | + if err != nil { |
| 211 | + return nil, err |
| 212 | + } |
| 213 | + sb.WriteString(fmt.Sprintf(" \"patterns\": %s\n", patternsJSON)) |
| 214 | + sb.WriteString("}\n") |
| 215 | + |
| 216 | + return []byte(sb.String()), nil |
| 217 | +} |
| 218 | + |
| 219 | +func fatal(format string, args ...any) { |
| 220 | + fmt.Fprintf(os.Stderr, "gen_inference_defaults: "+format+"\n", args...) |
| 221 | + os.Exit(1) |
| 222 | +} |
0 commit comments